Author: admin

What is a service robot? Explanation of definitions and usage examples
Today, due to the declining birthrate and aging population, the shortage of human resources and the aging of the workforce are accelerating, and there is a need to both save labor and improve productivity. The emphasis here is on the use of service robots. This article provides an overview of service robots, the benefits of introducing them, and specific usage scenarios.

What is a service robot? Explain the definition
Service robot are robots that are used in public places, medical facilities, stores, hotels, commercial facilities, and general households. The main purpose is to improve the quality of services and comfort of life, and it is used in a variety of areas such as transportation, reception, guidance, cleaning, nursing care, catering, security, inspection, and inventory management. In recent years, many office buildings have introduced complex service robots that are equipped with multiple functions such as security, reception, and guidance.
Difference between service robots and industrial robots
We are currently at the dawn of the Fourth Industrial Revolution, and robotics is attracting attention as one of the technologies that will support its realization. The robot market can be broadly divided into service robots and industrial robots. These differences are where the technology and features are utilized. The Japanese Industrial Standards define the difference between service robots and industrial robots as follows:
■Industrial robots
A robot that is an automatically controlled, reprogrammable, versatile manipulator, programmable in three or more axes, fixed in place or with mobile capabilities, used in industrial automation applications.
■Service robot
A robot that performs tasks that are beneficial to people or equipment. Excludes those used for industrial automation purposes.
Industrial robots are capable of complex control of multi-axis manipulators (arm-shaped remote control devices), and are primarily used to replace human work at manufacturing sites. Service robots, on the other hand, support human tasks and movements for purposes other than industrial use. In other words, industrial robots are robots that are mainly used in production factories and production lines, and service robots are robots that are used in other areas of life and services. Additionally, unlike industrial robots that simply carry out predetermined actions, service robots are characterized by the fact that many of them can communicate with humans. Furthermore, service robots are broadly divided into commercial and personal use, and indoor and outdoor use.
Advantages of introducing service robots
There are three main benefits of using service robots in business fields:
Can reduce labor costs
By introducing service robots, it is possible to reduce labor costs. For example, serving robots are becoming popular in the food and beverage industry. AI-equipped robots automatically serve food and drinks, significantly reducing the workload on hall staff. In this way, the introduction of service robots not only contributes to reducing labor costs, but also eliminates imbalances in hourly wages such as late at night, weekends, and holidays when there are fewer workers.
In Japan, as the shortage of human resources worsens due to the effects of the declining birthrate and aging population, the average hourly wage of part-time workers is increasing year by year. Therefore, how to reduce labor costs while maintaining service quality and labor productivity is an important issue for small and medium-sized enterprises and privately run restaurants that lack financial resources. Although introducing a service robot requires a certain amount of cost, it can reduce labor costs by reducing the workload and contribute to stabilizing cash flow.
Improved customer satisfaction
The introduction of service robots will also lead to improved customer satisfaction. For example, if tasks such as serving and preparing meals at a restaurant can be automated, hall staff will have more free time to focus on customer service. This makes it possible to provide personalized and attentive service to each customer, which can be expected to improve overall customer satisfaction.
Another major benefit is that if service robots can replace tasks such as office reception and cleaning, freed human resources can be focused on core tasks that directly lead to improved business performance. Originally, there is no superiority or inferiority to any business, but since a company’s management resources are limited, it is extremely important to distinguish between core business and non-core business. Concentrating resources on highly important tasks will contribute to improving the quality of products and services, which in turn will lead to maximizing customer satisfaction.
Eliminating labor shortages
According to data from the Statistics Bureau of the Ministry of Internal Affairs and Communications, the estimated total population of Japan is 124.56 million as of July 1, 2023, which has continued to decline since peaking at 128.08 million in 2008. As a result, the working-age population is decreasing, and the current situation in Japan is that the shortage of human resources and the aging of the workforce are becoming increasingly serious in various fields. In order for companies to continue to develop in this social context, they must both reduce labor and improve productivity.
If non-core tasks can be streamlined and automated by introducing service robots, tasks that would previously require multiple people can now be handled by a small number of people, making it possible to achieve productivity equal to or higher than before with fewer resources. Additionally, in order to design the placement and wiring of service robots, it is necessary to understand the overall picture of the business process. If the existing business flow can be visualized during this process, it has the advantage of contributing to streamlining operations and reducing man-hours.

Service robot introduction example
Here, we will introduce scenarios in which “security robots” and “cleaning robots” are used as typical examples of service robots.
Security robot example
One of the typical applications of service robots is security work for office buildings, commercial facilities, etc. Robots equipped with IoT sensors and cameras autonomously guard the building and detect suspicious movements and abnormalities. Among these, security robots are particularly good at patrolling within facilities and monitoring (sentry) at specific locations. For example, one office building had six security guards guarding 20 floors. However, there is a case in which the introduction of four security robots automated patrolling and monitoring, cutting the number of security guards in half to three.
Example of cleaning robot
Cleaning robots are service robots that are increasingly being introduced in public facilities, commercial facilities, medical facilities, restaurants, etc. It can not only clean floors, walls, windows, etc., but also sterilize them. Since robots automatically run and clean the facility, there is less unevenness in quality, and it is also possible to clean and sterilize places that humans cannot reach or dangerous areas. Coupled with the renewed recognition of the importance of disinfection and sterilization due to the spread of the new coronavirus, the number of cases in which robots are responsible for cleaning and sterilization work is increasing.

Utilization of robots in the office and future prospects
Service robots are a technology that is attracting attention in various fields, but their current scope of use is extremely limited, such as cleaning, security, reception, and guidance. However, information and communication technology is developing rapidly, and as AI and IoT become more sophisticated, the scope of its use is expected to expand, including improving the efficiency of backyard operations and managing the performance of human resources.
For example, service robots could be used to manage employee health to help maximize performance, or to support communication by translating and summarizing languages. In the future, it will become an intermediate platform that connects the office environment and online environment, and may become an indispensable solution for building a digital workplace.

Summary
Service robots are robots whose purpose is to improve the quality of life and services. The introduction of service robots reduces the work load on human resources, providing benefits such as “reducing personnel costs,” “improving customer satisfaction,” and “resolving human resource shortages.” It continues to attract a lot of attention as a technology that will be used in a variety of office environments.

Follow us on Facebook for updates and exclusive content! Click here: Each Techy
June 20, 2024
What is an AI algorithm?
AI algorithms are constantly advancing, and new papers and services are being published every day.
On the other hand, some systems that are currently called “AI” actually run on classical algorithms. Many of them are based on old-fashioned statistical methods.
In this article, I will list representative algorithms of AI and introduce the basic algorithm mechanism.

What is an AI algorithm?
The word ” algorithm ” is difficult to describe in one sentence. Above all, there is also no proper translation in Japanese.
When the word ” algorithm ” is commonly used, it refers to some procedure. Algorithm in AI is also close to its understanding, and in short, it refers to “calculation procedure”.
Basic structure of the algorithm
The most common algorithm is “sort”. For example, let’s consider the problem “Sort the horizontally aligned numbers in ascending order”.
There are various ways to solve the problem and ways of thinking about it, but let’s simply check the magnitude relationship of the numbers from the left.
If the number on the right is smaller than the number on the left, the left-right relationship is flipped. Repeat this process until the number on the right is greater than the number on the left.

This procedure is the procedure of calculation, that is, the algorithm. Did you get an image of the algorithm?
Advantages of Algorithms
By implementing an algorithm programmatically, anyone with that program can use that algorithm.
Algorithms in AI are similar, and even if we don’t know how the algorithm works or how it is mathematically designed, we can get results just by using that algorithm. .
This is because algorithms implemented in programming are in the form of “functions”. A programmatic function is something that transforms input into output.
It’s perfectly fine for us to use the program without knowing how the functions work internally.
But understanding how algorithms work can give us a better understanding of AI itself. In this article, I will introduce various algorithms, but for the sake of intuitiveness, I will try to avoid mathematical explanations.
Algorithms for supervised learning
Regression and classification
Supervised learning methods can be broadly divided into regression and classification. Regression techniques deal with the problem of “predicting future numbers” for some data, while classification techniques deal with the problem of “predicting which class some data belongs to”.
In other words, regression techniques deal with ‘continuous values’, whereas classification techniques deal with ‘discrete values’. The figure below shows the difference between regression and classification.
regression analysis
Regression analysis predicts the target variable you want to predict based on various other explanatory variables.
When there is only one explanatory variable, it is called simple regression analysis. By interpreting the objective variable y as the dependent variable and the explanatory variable x as the independent variable, simple regression analysis can be expressed as a linear function of the form “y=ax+b” with a and b as parameters. When there are multiple explanatory variables, it is called multiple regression analysis.

There is a distinction between “linear regression” and “nonlinear regression” in regression analysis. This is an intuitive explanation that lacks rigor, but a regression analysis that can linearly express the relationship shown in the figure above, in other words, the relationship between data is called “linear regression.”
k-nearest neighbor method
A typical classification problem algorithm is “k-nearest neighbor”. It determines to which class unknown data belongs to class-divided data scattered on coordinates.
Extract k pieces of data from the unknown data in descending order of distance, and sort the unknown data into the class with the largest number among the k pieces of data. The diagram is as follows.
Determine to which class the unknown data belongs to the already labeled data group. In this example, there are three classes: the red circle class, the blue star class, and the green diamond class.
Next, with k=3, three data are extracted from the unknown data in descending order of distance. In this example, there are 1 blue star and 2 green diamonds, so a majority vote is taken to determine that this unknown data belongs to the green diamond class.
Random forest
A random forest is a combination of several algorithms called “ decision trees ”. It may be easier to understand what a decision tree is by expressing it in a flow chart as shown below.
The image above shows a decision tree with YES/NO answers to questions.
Random forest refers to an algorithm that arranges multiple decision trees and decides the result by majority vote.
Also, since there are two types of decision trees: regression decision trees and classification decision trees, random forests can handle both problems.
Support vector machine
A support vector machine is an algorithm that calculates “margin maximization” for a data group. Let’s follow the process with reference to the diagram.
Let’s consider the problem of separating red circles and blue stars from scattered data with a “boundary line”. However, as you can see in this figure, there are many ways to draw the line.
Now consider “maximizing the support vector margin”. Support vectors refer to the data near the border, and margin refers to the distance between the border and the data. The green line in the figure is the margin.
The line that maximizes this margin is taken as the boundary line. This way you can avoid “false positives”. This is because maximizing the margin reduces the number of data that are ambiguous as to which of the two classes they belong to.
This support vector machine is an algorithm that can be used for both regression and classification problems.
Algorithms for unsupervised learning
clustering
A typical unsupervised learning algorithm is ” clustering “.
Clustering is an algorithm for grouping unknown data. The difference from the so-called classification ( supervised learning ) algorithm can be expressed as shown in the figure below.
k-means method
The k-means method is the most commonly used clustering algorithm.
First, randomly determine k centroid points for the scattered data group and use them as the core.
Then, the distances to the k nuclei are calculated for all data and grouped into the closest nuclei. This group is called a “cluster”.
Next, find the center of gravity for each cluster and use it as the new k kernels. Repeat the same process to separate each data into the nearest centroid clusters.
Repeat this process until the center of mass no longer moves. The calculation ends when the centroid point is no longer updated.
Reinforcement learning algorithm
Q-learning
Q-learning is an “algorithm that learns the Q value”. Understanding mathematical formulas is an unavoidable part of learning Q-learning, but here I will try to simplify it as much as possible.
Q-learning can be expressed by the following formula.
This algorithm can be interpreted as “choose the action a that maximizes the reward r in the state s”.
The expected value of the reward that can be obtained by taking that action is expressed as the Q value. Since the current state s is created as a result of accumulating the value of past actions, the current state s always has a Q value. And you can update the Q value depending on what action you take next. Choosing the action with the highest Q value increases the chances of reaching the reward.
There are two types of parameters, α and γ. α is the “learning rate”, which determines how quickly the Q value is updated. γ is the “discount rate” and represents how much we can trust the Q-value of the next action to incorporate it into the current Q-value . Optimizing this will result in proper learning.
Other reinforcement learning algorithms
Other reinforcement learning algorithms include Monte Carlo and SARSA. The Monte Carlo method is a fairly classical algorithm, but it takes a long time to learn because the reward-seeking process cannot be sequential.
A reinforcement learning algorithm called TD learning overcomes this drawback, and SARSA belongs to the same TD learning algorithm as Q learning.
Summary
In this article, we introduced a typical AI algorithm. Understanding algorithms leads to understanding how Artificial Intelligence works.
The algorithms presented here are the most basic and only scratch the surface. It will be more advanced content, but if you are interested in the latest AI, it is a good idea to follow the trend of cutting-edge algorithms.
Interestingly, some classical AI algorithms have achieved great results by combining them with deep learning techniques. The mechanism of AI is still in the stage of fumbling, and you can see that it is ” not easy “.

Follow us on Facebook for updates and exclusive content! Click here: Each Techy
June 12, 2024
What is an AI algorithm?
AI algorithms are constantly advancing, and new papers and services are being published every day.
On the other hand, some systems that are currently called “AI” actually run on classical algorithms. Many of them are based on old-fashioned statistical methods.
In this article, I will list representative algorithms of AI and introduce the basic algorithm mechanism.

What is an AI algorithm?
The word ” algorithm ” is difficult to describe in one sentence. Above all, there is also no proper translation in Japanese.
When the word ” algorithm ” is commonly used, it refers to some procedure. Algorithm in AI is also close to its understanding, and in short, it refers to “calculation procedure”.
Basic structure of the algorithm
The most common algorithm is “sort”. For example, let’s consider the problem “Sort the horizontally aligned numbers in ascending order”.
There are various ways to solve the problem and ways of thinking about it, but let’s simply check the magnitude relationship of the numbers from the left.
If the number on the right is smaller than the number on the left, the left-right relationship is flipped. Repeat this process until the number on the right is greater than the number on the left.

This procedure is the procedure of calculation, that is, the algorithm. Did you get an image of the algorithm?
Advantages of Algorithms
By implementing an algorithm programmatically, anyone with that program can use that algorithm.
Algorithms in AI are similar, and even if we don’t know how the algorithm works or how it is mathematically designed, we can get results just by using that algorithm. .
This is because algorithms implemented in programming are in the form of “functions”. A programmatic function is something that transforms input into output.
It’s perfectly fine for us to use the program without knowing how the functions work internally.
But understanding how algorithms work can give us a better understanding of AI itself. In this article, I will introduce various algorithms, but for the sake of intuitiveness, I will try to avoid mathematical explanations.
Algorithms for supervised learning
Regression and classification
Supervised learning methods can be broadly divided into regression and classification. Regression techniques deal with the problem of “predicting future numbers” for some data, while classification techniques deal with the problem of “predicting which class some data belongs to”.
In other words, regression techniques deal with ‘continuous values’, whereas classification techniques deal with ‘discrete values’. The figure below shows the difference between regression and classification.
regression analysis
Regression analysis predicts the target variable you want to predict based on various other explanatory variables.
When there is only one explanatory variable, it is called simple regression analysis. By interpreting the objective variable y as the dependent variable and the explanatory variable x as the independent variable, simple regression analysis can be expressed as a linear function of the form “y=ax+b” with a and b as parameters. When there are multiple explanatory variables, it is called multiple regression analysis.

There is a distinction between “linear regression” and “nonlinear regression” in regression analysis. This is an intuitive explanation that lacks rigor, but a regression analysis that can linearly express the relationship shown in the figure above, in other words, the relationship between data is called “linear regression.”
k-nearest neighbor method
A typical classification problem algorithm is “k-nearest neighbor”. It determines to which class unknown data belongs to class-divided data scattered on coordinates.
Extract k pieces of data from the unknown data in descending order of distance, and sort the unknown data into the class with the largest number among the k pieces of data. The diagram is as follows.
Determine to which class the unknown data belongs to the already labeled data group. In this example, there are three classes: the red circle class, the blue star class, and the green diamond class.
Next, with k=3, three data are extracted from the unknown data in descending order of distance. In this example, there are 1 blue star and 2 green diamonds, so a majority vote is taken to determine that this unknown data belongs to the green diamond class.
Random forest
A random forest is a combination of several algorithms called “ decision trees ”. It may be easier to understand what a decision tree is by expressing it in a flow chart as shown below.
The image above shows a decision tree with YES/NO answers to questions.
Random forest refers to an algorithm that arranges multiple decision trees and decides the result by majority vote.
Also, since there are two types of decision trees: regression decision trees and classification decision trees, random forests can handle both problems.
Support vector machine
A support vector machine is an algorithm that calculates “margin maximization” for a data group. Let’s follow the process with reference to the diagram.
Let’s consider the problem of separating red circles and blue stars from scattered data with a “boundary line”. However, as you can see in this figure, there are many ways to draw the line.
Now consider “maximizing the support vector margin”. Support vectors refer to the data near the border, and margin refers to the distance between the border and the data. The green line in the figure is the margin.
The line that maximizes this margin is taken as the boundary line. This way you can avoid “false positives”. This is because maximizing the margin reduces the number of data that are ambiguous as to which of the two classes they belong to.
This support vector machine is an algorithm that can be used for both regression and classification problems.
Algorithms for unsupervised learning
clustering
A typical unsupervised learning algorithm is ” clustering “.
Clustering is an algorithm for grouping unknown data. The difference from the so-called classification ( supervised learning ) algorithm can be expressed as shown in the figure below.
k-means method
The k-means method is the most commonly used clustering algorithm.
First, randomly determine k centroid points for the scattered data group and use them as the core.
Then, the distances to the k nuclei are calculated for all data and grouped into the closest nuclei. This group is called a “cluster”.
Next, find the center of gravity for each cluster and use it as the new k kernels. Repeat the same process to separate each data into the nearest centroid clusters.
Repeat this process until the center of mass no longer moves. The calculation ends when the centroid point is no longer updated.
Reinforcement learning algorithm
Q-learning
Q-learning is an “algorithm that learns the Q value”. Understanding mathematical formulas is an unavoidable part of learning Q-learning, but here I will try to simplify it as much as possible.
Q-learning can be expressed by the following formula.
This algorithm can be interpreted as “choose the action a that maximizes the reward r in the state s”.
The expected value of the reward that can be obtained by taking that action is expressed as the Q value. Since the current state s is created as a result of accumulating the value of past actions, the current state s always has a Q value. And you can update the Q value depending on what action you take next. Choosing the action with the highest Q value increases the chances of reaching the reward.
There are two types of parameters, α and γ. α is the “learning rate”, which determines how quickly the Q value is updated. γ is the “discount rate” and represents how much we can trust the Q-value of the next action to incorporate it into the current Q-value . Optimizing this will result in proper learning.
Other reinforcement learning algorithms
Other reinforcement learning algorithms include Monte Carlo and SARSA. The Monte Carlo method is a fairly classical algorithm, but it takes a long time to learn because the reward-seeking process cannot be sequential.
A reinforcement learning algorithm called TD learning overcomes this drawback, and SARSA belongs to the same TD learning algorithm as Q learning.
Summary
In this article, we introduced a typical AI algorithm. Understanding algorithms leads to understanding how Artificial Intelligence works.
The algorithms presented here are the most basic and only scratch the surface. It will be more advanced content, but if you are interested in the latest AI, it is a good idea to follow the trend of cutting-edge algorithms.
Interestingly, some classical AI algorithms have achieved great results by combining them with deep learning techniques. The mechanism of AI is still in the stage of fumbling, and you can see that it is ” not easy “.

Follow us on Facebook for updates and exclusive content! Click here: Each Techy
June 12, 2024
What is an AI algorithm?
AI algorithms are constantly advancing, and new papers and services are being published every day.
On the other hand, some systems that are currently called “AI” actually run on classical algorithms. Many of them are based on old-fashioned statistical methods.
In this article, I will list representative algorithms of AI and introduce the basic algorithm mechanism.

What is an AI algorithm?
The word ” algorithm ” is difficult to describe in one sentence. Above all, there is also no proper translation in Japanese.
When the word ” algorithm ” is commonly used, it refers to some procedure. Algorithm in AI is also close to its understanding, and in short, it refers to “calculation procedure”.
Basic structure of the algorithm
The most common algorithm is “sort”. For example, let’s consider the problem “Sort the horizontally aligned numbers in ascending order”.
There are various ways to solve the problem and ways of thinking about it, but let’s simply check the magnitude relationship of the numbers from the left.
If the number on the right is smaller than the number on the left, the left-right relationship is flipped. Repeat this process until the number on the right is greater than the number on the left.

This procedure is the procedure of calculation, that is, the algorithm. Did you get an image of the algorithm?
Advantages of Algorithms
By implementing an algorithm programmatically, anyone with that program can use that algorithm.
Algorithms in AI are similar, and even if we don’t know how the algorithm works or how it is mathematically designed, we can get results just by using that algorithm. .
This is because algorithms implemented in programming are in the form of “functions”. A programmatic function is something that transforms input into output.
It’s perfectly fine for us to use the program without knowing how the functions work internally.
But understanding how algorithms work can give us a better understanding of AI itself. In this article, I will introduce various algorithms, but for the sake of intuitiveness, I will try to avoid mathematical explanations.
Algorithms for supervised learning
Regression and classification
Supervised learning methods can be broadly divided into regression and classification. Regression techniques deal with the problem of “predicting future numbers” for some data, while classification techniques deal with the problem of “predicting which class some data belongs to”.
In other words, regression techniques deal with ‘continuous values’, whereas classification techniques deal with ‘discrete values’. The figure below shows the difference between regression and classification.
regression analysis
Regression analysis predicts the target variable you want to predict based on various other explanatory variables.
When there is only one explanatory variable, it is called simple regression analysis. By interpreting the objective variable y as the dependent variable and the explanatory variable x as the independent variable, simple regression analysis can be expressed as a linear function of the form “y=ax+b” with a and b as parameters. When there are multiple explanatory variables, it is called multiple regression analysis.

There is a distinction between “linear regression” and “nonlinear regression” in regression analysis. This is an intuitive explanation that lacks rigor, but a regression analysis that can linearly express the relationship shown in the figure above, in other words, the relationship between data is called “linear regression.”
k-nearest neighbor method
A typical classification problem algorithm is “k-nearest neighbor”. It determines to which class unknown data belongs to class-divided data scattered on coordinates.
Extract k pieces of data from the unknown data in descending order of distance, and sort the unknown data into the class with the largest number among the k pieces of data. The diagram is as follows.
Determine to which class the unknown data belongs to the already labeled data group. In this example, there are three classes: the red circle class, the blue star class, and the green diamond class.
Next, with k=3, three data are extracted from the unknown data in descending order of distance. In this example, there are 1 blue star and 2 green diamonds, so a majority vote is taken to determine that this unknown data belongs to the green diamond class.
Random forest
A random forest is a combination of several algorithms called “ decision trees ”. It may be easier to understand what a decision tree is by expressing it in a flow chart as shown below.
The image above shows a decision tree with YES/NO answers to questions.
Random forest refers to an algorithm that arranges multiple decision trees and decides the result by majority vote.
Also, since there are two types of decision trees: regression decision trees and classification decision trees, random forests can handle both problems.
Support vector machine
A support vector machine is an algorithm that calculates “margin maximization” for a data group. Let’s follow the process with reference to the diagram.
Let’s consider the problem of separating red circles and blue stars from scattered data with a “boundary line”. However, as you can see in this figure, there are many ways to draw the line.
Now consider “maximizing the support vector margin”. Support vectors refer to the data near the border, and margin refers to the distance between the border and the data. The green line in the figure is the margin.
The line that maximizes this margin is taken as the boundary line. This way you can avoid “false positives”. This is because maximizing the margin reduces the number of data that are ambiguous as to which of the two classes they belong to.
This support vector machine is an algorithm that can be used for both regression and classification problems.
Algorithms for unsupervised learning
clustering
A typical unsupervised learning algorithm is ” clustering “.
Clustering is an algorithm for grouping unknown data. The difference from the so-called classification ( supervised learning ) algorithm can be expressed as shown in the figure below.
k-means method
The k-means method is the most commonly used clustering algorithm.
First, randomly determine k centroid points for the scattered data group and use them as the core.
Then, the distances to the k nuclei are calculated for all data and grouped into the closest nuclei. This group is called a “cluster”.
Next, find the center of gravity for each cluster and use it as the new k kernels. Repeat the same process to separate each data into the nearest centroid clusters.
Repeat this process until the center of mass no longer moves. The calculation ends when the centroid point is no longer updated.
Reinforcement learning algorithm
Q-learning
Q-learning is an “algorithm that learns the Q value”. Understanding mathematical formulas is an unavoidable part of learning Q-learning, but here I will try to simplify it as much as possible.
Q-learning can be expressed by the following formula.
This algorithm can be interpreted as “choose the action a that maximizes the reward r in the state s”.
The expected value of the reward that can be obtained by taking that action is expressed as the Q value. Since the current state s is created as a result of accumulating the value of past actions, the current state s always has a Q value. And you can update the Q value depending on what action you take next. Choosing the action with the highest Q value increases the chances of reaching the reward.
There are two types of parameters, α and γ. α is the “learning rate”, which determines how quickly the Q value is updated. γ is the “discount rate” and represents how much we can trust the Q-value of the next action to incorporate it into the current Q-value . Optimizing this will result in proper learning.
Other reinforcement learning algorithms
Other reinforcement learning algorithms include Monte Carlo and SARSA. The Monte Carlo method is a fairly classical algorithm, but it takes a long time to learn because the reward-seeking process cannot be sequential.
A reinforcement learning algorithm called TD learning overcomes this drawback, and SARSA belongs to the same TD learning algorithm as Q learning.
Summary
In this article, we introduced a typical AI algorithm. Understanding algorithms leads to understanding how Artificial Intelligence works.
The algorithms presented here are the most basic and only scratch the surface. It will be more advanced content, but if you are interested in the latest AI, it is a good idea to follow the trend of cutting-edge algorithms.
Interestingly, some classical AI algorithms have achieved great results by combining them with deep learning techniques. The mechanism of AI is still in the stage of fumbling, and you can see that it is ” not easy “.

Follow us on Facebook for updates and exclusive content! Click here: Each Techy
June 12, 2024
How to Choose a freelance online bank! 3 Recommended Bank Accounts
A freelance engineer will explain why you should choose an online bank for your business account, the benefits, and how to choose a bank. We will introduce 3 recommended banks for you while comparing interest rates, various transaction fees, and profitable conditions. Choose the bank that suits your lifestyle wisely and manage your money!
1. Benefits of freelance choosing an online bank
If you are thinking of working as a freelancer (individual business owner), or if you are currently managing your personal account as your main account, we recommend that you create a bank account for your business. To do.
There are three main types of banks, even if they are called bank accounts.
Online bank
If you work freelance, we recommend online banking.
Unlike salaries, who receive salaries from the company once a month, in the case of freelancers, the timing of payment of compensation depends on the client. If you use an online bank, you can check it from your computer or smartphone regardless of location or time, so it will not be a hassle.
In addition, in the case of online banks, it is often possible to link with cloud accounting software, which has many advantages for freelancers, such as making it easier to book double-entry bookkeeping for blue tax returns, which is required when filing tax returns.

[Main benefits of online banks]
・ Fees are relatively low
・ Deposit interest rates are relatively high ・ Deposits
and withdrawals can be easily made at affiliated ATMs ・
Time can be saved because online procedures can be completed ・ Procedures can be made
24 hours a day, 365 days a year
・ Bookkeeping You don’t have to go out and you can download transaction records to your computer.
・ You can check your balance at any time on your smartphone. ・
There is no account maintenance fee.
・ You can often link with cloud accounting software.
Mega bank
It is well known and can give trust to trading partners.
In addition, there are ATMs and branches all over the country, and it is attractive that it is highly convenient, but the fact that the examination of opening an account is strict and the account maintenance fee is relatively high may be a disadvantage for freelancers. ..
Shinkin bank
As a community-based financial institution, Shinkin banks have a mission to support the local economy, mainly dealing with sole proprietors and small and medium-sized enterprises, so even freelancers can easily borrow money.
2. Points for choosing an online bank
For those who are thinking of opening a business account, we would like to introduce the points that we would like to emphasize when choosing an online bank.
2.1 Choose a bank with a high interest rate
In the case of online banks, the interest rate of ordinary deposits is often high, so it is recommended to pay attention to the interest rate of ordinary deposits rather than fixed deposits. The interest rate on ordinary deposits is set at 0.001 to 0.15%, which is higher than the basic value of 0.001% for Mega Bank. As for fixed deposits, mega banks have a high interest rate of 0.01%, while online banks have a high interest rate of 0.02%.
2.2 Select by ATM fee & transfer fee
In most online banks, ATM fees and transfer fees to other banks are free with a limited number of times, depending on the rank, stage, and affiliated convenience store. Depending on the bank, the number of times it will be free varies, so check the conditions and choose a good bank.
2.3 Select a bank that can work with cloud accounting software
Recent cloud accounting software has a function to automatically capture the details of transactions made through an online bank and enter them in the books. The banks that work with your accounting software differ, so choose a bank that works with your accounting software.

◆ If you haven’t selected accounting software (tax return software) yet, please refer to this article as well.
Compare and select tax return software recommended for freelance engineers!
3. 3 recommended online banks for freelance
We will introduce carefully selected online banks, focusing on banks with high interest rates on ordinary deposits and banks with low ATM fees and transfer fees.
The conditions for raising interest rates and getting commissions vary from bank to bank. Depending on whether you have an ATM in your area or where you often shop, the banks you can use at a great price will change. We recommend that you choose a bank that is convenient for you in the light of your own life.
3.1 Popular with freelancers! Rakuten Bank is recommended for Rakuten users
A major feature of this service is that the interest rate on savings deposits can be significantly increased from 0.02% to 0.10% by using the “Money Bridge” service, which makes it convenient to transfer funds between accounts with Rakuten Securities. Even if you say that the accounts are linked, you do not need to invest, and it will be applied if you open each account under the same name.
In addition, Rakuten points can be earned by trading with other bank accounts or using ATMs, so it is a recommended bank with many benefits for Rakuten users.
It is also a nice point for freelancers to be able to open an account with a store name, which is rare in online banks.
・ Ordinary deposit interest rate: 0.02 to 0.10% ( * When using “Money Bridge” of “Rakuten Securities” )
・ Fixed deposit interest rate: 0.02%
・ ATM fee: Available at approximately 100,000 ATMs nationwide

Convenience ATM withdrawal fee
Seven-Eleven Lawson FamilyMart
(E-net) Ministop
(AEON Bank)
Free up to 7 times a month

・ Transfer fee: Free up to 3 times a month
* The number of free ATM fees and transfer fees varies depending on the rank of “Happy Program”.
・ Opening an account with a store name: Yes (You can add a store name to your personal name. Example: Name + store name, store name + name)
3.2 Maximum interest rate 0.10%! Completely free fees using AEON Bank ATMs
If AEON Bank meets simple conditions such as signing a credit card “AEON Select Card” with no annual membership fee, registering for “AEON Bank” Internet banking, and using WAON Auto Charge, “AEON Bank My Stage” will be displayed. “Silver Stage” will be applied and the interest rate on savings deposits will increase to 0.03%. In addition, the AEON Bank score will be given according to the amount of WAON used, the stage will rise, and the interest rate will rise to a maximum of 0.10%.
If you use more than 6,000 ATMs of AEON Bank nationwide, the fee will be completely free 24 hours a day, 365 days a year. For people , AEON Bank is recommended.
・ Ordinary deposit interest rate: 0.01-0.10% ( * In the case of “Platinum stage” of “AEON Bank My Stage” )
・ Fixed deposit interest rate: 0.02%
・ ATM fee: 24 hours a day, 365 days a year Over 6,000 units. There are approximately 55,000 affiliated financial institution ATMs nationwide.

Convenience ATM withdrawal fee
Seven-Eleven Lawson FamilyMart
(E-net) Ministop
(AEON Bank)
No handling Free up to 5 times a month Unlimited 24 hours a day, 365 days a year

・ Transfer fee: Free up to 5 times a month
* The number of free ATM fees and transfer fees varies depending on the stage of “AEON Bank My Stage”.
・ Opening an account with a store name: Not possible
3.3 Great deals on various transaction fees! Sumishin SBI Net Bank
The main point of SBI Sumishin Net Bank is that various transaction fees are set at a reasonable price. At SBI Sumishin Net Bank, the number of free ATM withdrawal fees and transfer fees varies depending on the rank of the smart program, but the lowest rank is twice a month, the maximum is 15 times a month, and the ATM withdrawal fee is free. increase. Similarly, the transfer fee to other banks is free up to 15 times a month, and even after the number of free transfers, it is set at 157 yen, which is cheaper than other online banks. Depending on the rank, SBI Sumishin Net Bank is recommended if you want to save various transaction fees that
are not surprisingly stupid .
・ Ordinary deposit interest rate: 0.001 to 0.01% ( * When “hybrid deposit” linked to SBI securities account is applied )
・ Fixed deposit interest rate: 0.02%
・ ATM fee: 24 hours a day, 365 days a year from nearby ATMs such as convenience stores and supermarkets nationwide * Transactions are possible
* Transactions are restricted for a few minutes (due to system processing) every Saturday from 24:00 (0:00 on Sunday) during system maintenance hours of our company and partner ATMs.

Convenience ATM withdrawal fee
Seven-Eleven Lawson FamilyMart
(E-net) Ministop
(AEON Bank)
Free from 2 times a month up to 15 times

・ Transfer fee: Free once to 15 times a month
* The number of free ATM fees and transfer fees varies depending on the rank of the “Smart Program”.
・ Opening an account with a store name: Not possible
4. Why it’s better to have a business account for freelance
For those who work as a freelancer (sole proprietor), we strongly recommend that you separate your personal account from your business account. There are three main reasons.

1. Makes it easier to understand the financial situation
If your personal and business accounts are the same, it’s hard to tell if your business is profitable or in the red because your living and business spending are mixed.
It is difficult to separate the income and expenditure for daily life and the expenditure for business, so let’s manage the personal account and the business account clearly to make it easy to understand the inflow and outflow of money for business purposes.

2. Bookkeeping becomes smooth
If you want to start your business as a freelancer, you will need to book the double-entry bookkeeping of the blue tax return in the final tax return. At this time, if you have a business account, you can mechanically drop the contents recorded in the account into the bookkeeping, so bookkeeping will be smooth.
Also, depending on the online bank and accounting software you use, the transaction details will be automatically imported and entered in the bookkeeping, making it easier to create bookkeeping.

3. Effective for tax audit measures
For freelancers, this “tax audit measure” is of utmost importance. A tax audit is the process of investigating whether a tax return is correct and whether there is tax evasion.
Freelancers should be especially careful about confusing personal and business accounts, and there is a great risk that personal shopping will be expensed and pointed out. To prevent such accidental mistakes, manage your accounts separately.

10 freelance engineer tax-saving measures ranking! Save tax wisely with expenses and deductions
5. Let’s open a store name account
If you start a business as a freelance (individual business owner) and have an account for business, I think that there are many people who want to open an account with the name of the shop.
It has the advantage of gaining the trust and security of the trading partner, and you may consider opening an account with a store name to increase your motivation for work.
When opening a store name account, there are restrictions such as accepting only at the counter at Mega Bank, etc., or opening only the account of the branch office closest to your home or office, but with an online bank, from your home or office You can complete the procedure.
Although it depends on the financial institution, the documents generally required to open a store name account are as follows.
・ Identity verification documents
(driver’s license, my number card, etc.)
・ Business opening notification
・ Seal
・ Documents that can confirm the shop
name (rental contract with the shop name, utility bill receipt, business card, brochure, etc.)
For details, please contact each financial institution that opens an account.

Do freelancers need a business start notification? When will you release it? Is there any merit?
6. Summary
Freelancers often use a bank account to pay expenses or transfer rewards from clients. In order to clarify the business balance, why not consider opening a “business account” as your main account?
In that case, carefully check the conditions that can be used at a good price, and select a bank that offers convenient services that suit your lifestyle.
Opening a business account will increase your motivation and sense of responsibility for the job of “doing business as a freelancer.”
Follow us on Facebook for updates and exclusive content! Click here: Each Techy
June 11, 2024
How to Choose a freelance online bank! 3 Recommended Bank Accounts
A freelance engineer will explain why you should choose an online bank for your business account, the benefits, and how to choose a bank. We will introduce 3 recommended banks for you while comparing interest rates, various transaction fees, and profitable conditions. Choose the bank that suits your lifestyle wisely and manage your money!
1. Benefits of freelance choosing an online bank
If you are thinking of working as a freelancer (individual business owner), or if you are currently managing your personal account as your main account, we recommend that you create a bank account for your business. To do.
There are three main types of banks, even if they are called bank accounts.
Online bank
If you work freelance, we recommend online banking.
Unlike salaries, who receive salaries from the company once a month, in the case of freelancers, the timing of payment of compensation depends on the client. If you use an online bank, you can check it from your computer or smartphone regardless of location or time, so it will not be a hassle.
In addition, in the case of online banks, it is often possible to link with cloud accounting software, which has many advantages for freelancers, such as making it easier to book double-entry bookkeeping for blue tax returns, which is required when filing tax returns.

[Main benefits of online banks]
・ Fees are relatively low
・ Deposit interest rates are relatively high ・ Deposits
and withdrawals can be easily made at affiliated ATMs ・
Time can be saved because online procedures can be completed ・ Procedures can be made
24 hours a day, 365 days a year
・ Bookkeeping You don’t have to go out and you can download transaction records to your computer.
・ You can check your balance at any time on your smartphone. ・
There is no account maintenance fee.
・ You can often link with cloud accounting software.
Mega bank
It is well known and can give trust to trading partners.
In addition, there are ATMs and branches all over the country, and it is attractive that it is highly convenient, but the fact that the examination of opening an account is strict and the account maintenance fee is relatively high may be a disadvantage for freelancers. ..
Shinkin bank
As a community-based financial institution, Shinkin banks have a mission to support the local economy, mainly dealing with sole proprietors and small and medium-sized enterprises, so even freelancers can easily borrow money.
2. Points for choosing an online bank
For those who are thinking of opening a business account, we would like to introduce the points that we would like to emphasize when choosing an online bank.
2.1 Choose a bank with a high interest rate
In the case of online banks, the interest rate of ordinary deposits is often high, so it is recommended to pay attention to the interest rate of ordinary deposits rather than fixed deposits. The interest rate on ordinary deposits is set at 0.001 to 0.15%, which is higher than the basic value of 0.001% for Mega Bank. As for fixed deposits, mega banks have a high interest rate of 0.01%, while online banks have a high interest rate of 0.02%.
2.2 Select by ATM fee & transfer fee
In most online banks, ATM fees and transfer fees to other banks are free with a limited number of times, depending on the rank, stage, and affiliated convenience store. Depending on the bank, the number of times it will be free varies, so check the conditions and choose a good bank.
2.3 Select a bank that can work with cloud accounting software
Recent cloud accounting software has a function to automatically capture the details of transactions made through an online bank and enter them in the books. The banks that work with your accounting software differ, so choose a bank that works with your accounting software.

◆ If you haven’t selected accounting software (tax return software) yet, please refer to this article as well.
Compare and select tax return software recommended for freelance engineers!
3. 3 recommended online banks for freelance
We will introduce carefully selected online banks, focusing on banks with high interest rates on ordinary deposits and banks with low ATM fees and transfer fees.
The conditions for raising interest rates and getting commissions vary from bank to bank. Depending on whether you have an ATM in your area or where you often shop, the banks you can use at a great price will change. We recommend that you choose a bank that is convenient for you in the light of your own life.
3.1 Popular with freelancers! Rakuten Bank is recommended for Rakuten users
A major feature of this service is that the interest rate on savings deposits can be significantly increased from 0.02% to 0.10% by using the “Money Bridge” service, which makes it convenient to transfer funds between accounts with Rakuten Securities. Even if you say that the accounts are linked, you do not need to invest, and it will be applied if you open each account under the same name.
In addition, Rakuten points can be earned by trading with other bank accounts or using ATMs, so it is a recommended bank with many benefits for Rakuten users.
It is also a nice point for freelancers to be able to open an account with a store name, which is rare in online banks.
・ Ordinary deposit interest rate: 0.02 to 0.10% ( * When using “Money Bridge” of “Rakuten Securities” )
・ Fixed deposit interest rate: 0.02%
・ ATM fee: Available at approximately 100,000 ATMs nationwide

Convenience ATM withdrawal fee
Seven-Eleven Lawson FamilyMart
(E-net) Ministop
(AEON Bank)
Free up to 7 times a month

・ Transfer fee: Free up to 3 times a month
* The number of free ATM fees and transfer fees varies depending on the rank of “Happy Program”.
・ Opening an account with a store name: Yes (You can add a store name to your personal name. Example: Name + store name, store name + name)
3.2 Maximum interest rate 0.10%! Completely free fees using AEON Bank ATMs
If AEON Bank meets simple conditions such as signing a credit card “AEON Select Card” with no annual membership fee, registering for “AEON Bank” Internet banking, and using WAON Auto Charge, “AEON Bank My Stage” will be displayed. “Silver Stage” will be applied and the interest rate on savings deposits will increase to 0.03%. In addition, the AEON Bank score will be given according to the amount of WAON used, the stage will rise, and the interest rate will rise to a maximum of 0.10%.
If you use more than 6,000 ATMs of AEON Bank nationwide, the fee will be completely free 24 hours a day, 365 days a year. For people , AEON Bank is recommended.
・ Ordinary deposit interest rate: 0.01-0.10% ( * In the case of “Platinum stage” of “AEON Bank My Stage” )
・ Fixed deposit interest rate: 0.02%
・ ATM fee: 24 hours a day, 365 days a year Over 6,000 units. There are approximately 55,000 affiliated financial institution ATMs nationwide.

Convenience ATM withdrawal fee
Seven-Eleven Lawson FamilyMart
(E-net) Ministop
(AEON Bank)
No handling Free up to 5 times a month Unlimited 24 hours a day, 365 days a year

・ Transfer fee: Free up to 5 times a month
* The number of free ATM fees and transfer fees varies depending on the stage of “AEON Bank My Stage”.
・ Opening an account with a store name: Not possible
3.3 Great deals on various transaction fees! Sumishin SBI Net Bank
The main point of SBI Sumishin Net Bank is that various transaction fees are set at a reasonable price. At SBI Sumishin Net Bank, the number of free ATM withdrawal fees and transfer fees varies depending on the rank of the smart program, but the lowest rank is twice a month, the maximum is 15 times a month, and the ATM withdrawal fee is free. increase. Similarly, the transfer fee to other banks is free up to 15 times a month, and even after the number of free transfers, it is set at 157 yen, which is cheaper than other online banks. Depending on the rank, SBI Sumishin Net Bank is recommended if you want to save various transaction fees that
are not surprisingly stupid .
・ Ordinary deposit interest rate: 0.001 to 0.01% ( * When “hybrid deposit” linked to SBI securities account is applied )
・ Fixed deposit interest rate: 0.02%
・ ATM fee: 24 hours a day, 365 days a year from nearby ATMs such as convenience stores and supermarkets nationwide * Transactions are possible
* Transactions are restricted for a few minutes (due to system processing) every Saturday from 24:00 (0:00 on Sunday) during system maintenance hours of our company and partner ATMs.

Convenience ATM withdrawal fee
Seven-Eleven Lawson FamilyMart
(E-net) Ministop
(AEON Bank)
Free from 2 times a month up to 15 times

・ Transfer fee: Free once to 15 times a month
* The number of free ATM fees and transfer fees varies depending on the rank of the “Smart Program”.
・ Opening an account with a store name: Not possible
4. Why it’s better to have a business account for freelance
For those who work as a freelancer (sole proprietor), we strongly recommend that you separate your personal account from your business account. There are three main reasons.

1. Makes it easier to understand the financial situation
If your personal and business accounts are the same, it’s hard to tell if your business is profitable or in the red because your living and business spending are mixed.
It is difficult to separate the income and expenditure for daily life and the expenditure for business, so let’s manage the personal account and the business account clearly to make it easy to understand the inflow and outflow of money for business purposes.

2. Bookkeeping becomes smooth
If you want to start your business as a freelancer, you will need to book the double-entry bookkeeping of the blue tax return in the final tax return. At this time, if you have a business account, you can mechanically drop the contents recorded in the account into the bookkeeping, so bookkeeping will be smooth.
Also, depending on the online bank and accounting software you use, the transaction details will be automatically imported and entered in the bookkeeping, making it easier to create bookkeeping.

3. Effective for tax audit measures
For freelancers, this “tax audit measure” is of utmost importance. A tax audit is the process of investigating whether a tax return is correct and whether there is tax evasion.
Freelancers should be especially careful about confusing personal and business accounts, and there is a great risk that personal shopping will be expensed and pointed out. To prevent such accidental mistakes, manage your accounts separately.

10 freelance engineer tax-saving measures ranking! Save tax wisely with expenses and deductions
5. Let’s open a store name account
If you start a business as a freelance (individual business owner) and have an account for business, I think that there are many people who want to open an account with the name of the shop.
It has the advantage of gaining the trust and security of the trading partner, and you may consider opening an account with a store name to increase your motivation for work.
When opening a store name account, there are restrictions such as accepting only at the counter at Mega Bank, etc., or opening only the account of the branch office closest to your home or office, but with an online bank, from your home or office You can complete the procedure.
Although it depends on the financial institution, the documents generally required to open a store name account are as follows.
・ Identity verification documents
(driver’s license, my number card, etc.)
・ Business opening notification
・ Seal
・ Documents that can confirm the shop
name (rental contract with the shop name, utility bill receipt, business card, brochure, etc.)
For details, please contact each financial institution that opens an account.

Do freelancers need a business start notification? When will you release it? Is there any merit?
6. Summary
Freelancers often use a bank account to pay expenses or transfer rewards from clients. In order to clarify the business balance, why not consider opening a “business account” as your main account?
In that case, carefully check the conditions that can be used at a good price, and select a bank that offers convenient services that suit your lifestyle.
Opening a business account will increase your motivation and sense of responsibility for the job of “doing business as a freelancer.”
Follow us on Facebook for updates and exclusive content! Click here: Each Techy
June 11, 2024

Convenience ATM withdrawal fee
Seven-Eleven	Lawson	FamilyMart (E-net)	Ministop (AEON Bank)
Free up to 7 times a month

Convenience ATM withdrawal fee
Seven-Eleven	Lawson	FamilyMart (E-net)	Ministop (AEON Bank)
No handling	Free up to 5 times a month	Unlimited 24 hours a day, 365 days a year

Convenience ATM withdrawal fee
Seven-Eleven	Lawson	FamilyMart (E-net)	Ministop (AEON Bank)
Free from 2 times a month up to 15 times

Convenience ATM withdrawal fee
Seven-Eleven	Lawson	FamilyMart (E-net)	Ministop (AEON Bank)
Free up to 7 times a month

Convenience ATM withdrawal fee
Seven-Eleven	Lawson	FamilyMart (E-net)	Ministop (AEON Bank)
No handling	Free up to 5 times a month	Unlimited 24 hours a day, 365 days a year

Convenience ATM withdrawal fee
Seven-Eleven	Lawson	FamilyMart (E-net)	Ministop (AEON Bank)
Free from 2 times a month up to 15 times

3 Recommended PCs for machine learning / deep learning | Explain the necessary parts!

What is machine learning

Machine learning is one of the most important aspects of AI development, and is an effective method for predicting numerical values and identifying and classifying images.

One of the machine learning methods is ” neural network . A neural network is a technique inspired by the structure of the human brain and mimics the way neurons work.

Some of the neural networks include multi-layer perceptrons and deep learning.

What is deep learning?

Deep learning was developed to enhance the capabilities of the neural network mentioned in the previous section.

Deep learning is a neural network with a multi-layered structure, and is currently the mainstream of AI development.

The difference between machine learning and deep learning is that machine learning learns rules from data by itself. In deep learning, the computer itself learns the feature values that must be specified in machine learning.

To put it simply, machine learning involves specifying feature values by humans, and deep learning involves learning feature values as well.

Currently, deep learning is used for image recognition , speech processing, natural language processing , etc. a variety of situations around the world, including

From here, let’s see what kind of PC should be used when actually performing deep learning.

Parts required for deep learning PC

The following five parts are required for a PC for deep learning.

OS
CPU
GPUs
Memory
Storage

I will explain each.

OS

OS is an abbreviation of “Operating system”, and refers to the system software that controls the operation (operation, operation, and operation) of a computer. In terms of a PC, it is a system that connects the device and software of the PC and controls the device and software.

Major PC OSs include Microsoft’s Windows, Apple’s mac OS, and open source Linux.

In AI development, Windows is recommended because it is easy to expand functions, and Linux is also used for servers. For Windows, the pro series is better from the point of view of functions

CPU

CPU is an abbreviation for “Central Processing Unit” and is the central processing unit in a computer (the brain of the computer, so to speak). CPUs are versatile in their processing and can handle a variety of things.

When choosing a PC for deep learning, it is a good idea to choose a higher model from Intel’s CPU “core i5”.

GPUs

GPU is an abbreviation for “Graphics Processing Unit”, and is a computing device specialized for screen display and image processing such as 3D graphics.

GPUs are good at simple calculations and good at parallel processing, so they are a very important part in AI development.

GPU processing speed is several to 100 times faster than CPU processing speed, and GPU is essential for deep learning.

Memory

Memory is the temporary storage of your computer’s work. Since it is temporary, it is characterized by fast access so that the current contents can be retrieved immediately.

When choosing a PC for deep learning, it is a good idea to choose a memory of 16GB or more.

Storage

Storage, also known as “auxiliary storage”, stores data for a long period of time. What is called a hard disk or SSD is one of this storage.

There is no problem with the storage that is installed in a normal PC, but if it is 512 GB or more, it can handle large amounts of data, so you can rest assured.

Differences between deep learning PCs and ordinary PCs

There are three differences between deep learning PCs and ordinary PCs: differences in specifications for each part,''using Linux as the OS,” and “requiring a GPU.”

In addition, PCs for deep learning are a type of workstation, and feature higher performance than regular PCs.

Also, some people who are serious about deep learning development use a PC that they have assembled with the necessary parts themselves.

In the following, I will briefly introduce “ordinary PC”, “self-made PC”, and “workstation/deep learning PC” as a supplement.

Normal PC

Deep learning can be performed even on a PC that is normally sold if it is equipped with a GPU.

For those who are studying deep learning for the first time or who want to try deep learning, a normal PC may be fine.

Homemade PC

If you want to do full-scale AI development, you should use your own PC. We also recommend the BTO PC, which allows you to select parts to some extent.

BTO: An abbreviation for “Build To Order”, which means build-to-order manufacturing. Compared to commercially available finished PCs, you can freely customize the processor, memory, hard disk, mouse, storage, etc.

Workstation/PC for deep learning

Workstations are used by individuals for work such as CAD. If you find it difficult, remember that it is a version with good performance on a normal PC.

* CAD: Design support software for automobiles, architecture, and clothing.

Supplement: Server

In addition to the above three methods, there are other ways to develop on the server. A server is used by many users. For personal use, you should choose one of the above three options.

Should I make my own PC for deep learning?

Earlier, I mentioned that “Some people who are serious about deep learning development use their own PCs.”

Below, we will introduce the advantages and cautions of using a self-made PC for those who are wondering whether they should build their own PC for deep learning.

Advantages of using your own PC

The advantage of using a self-made PC is that it can be specialized for deep learning and machine learning.

Homemade PCs can be assembled to have higher specs than those sold at regular stores, so it is recommended when a server cannot be used.

Points to note when making your own

One thing to keep in mind when building your own PC for deep learning is that you cannot request guarantees or repairs from the sales company.

It goes without saying that you build your own PC, but basically if something goes wrong, you have to investigate and deal with it yourself, or pay a certain amount of money and ask for a PC repair.

Therefore, if you are not very familiar with PCs and machines, you need to be careful when building your own PC.

3 Recommended PCs for Deep Learning

From here, we will introduce recommended PCs for deep learning. The following three PCs are introduced this time.

DEEP-17FG102-i7K-VOXVI
THIRDWAVE Pro WORKSTATION X4612 standard model
HP ZBook Studio 15.6inch G8 Mobile Workstation new standard model

①DEEP-17FG102-i7K-VOXVI

machine learning

The first recommended PC is “DEEP-17FG102-i7K-VOXVI”.

OS	Ubuntu 18.04 LTS
CPU	Core i7-9700K Intel Z370
memory	DDR4-2400 SODIMM (PC4-19200) 16GB (8GB x 2)
Storage ①	250GB NVMe M.2 SSD
Storage②	1TB Serial-ATA HDD
drive	No optical drive
GPUs	GeForce RTX 2080 8GB GDDR6
display	17.3 type (matte color liquid crystal) full HD (1920 x 1080 dots)
price	32,3980 yen ~ (as of 2022/02/08)

It supports 8GB of high-speed GDDR6 memory similar to desktop and GPU Boost 4.0 that brings out GPU performance, so code created at the development site can be executed at a speed comparable to mobile environments.

Although this PC is a notebook PC, it boasts performance comparable to that of a desktop PC. It is one that can be used at the forefront of AI development, such as creating sample code, demonstrating, and giving presentations.

In addition, the same PC comes with ax Co., Ltd.’s demo software “ailia AI showcase”, so you can use various AI functions using trained models such as object detection, image classification, feature extraction, skeleton extraction, and personal identification. You can easily try it.

In addition, it supports the GPU Cloud platform “NGC (NVIDIA GPU Cloud)” that facilitates AI development, and the latest development environment can be used without complicated settings.

Just by downloading the deep learning framework, you can use it without worrying about complicated environment settings and consistency, so it is the best laptop for those who are just starting deep learning.

A framework is a piece of software that serves as the foundation upon which an application is developed.

②THIRDWAVE Pro WORKSTATION X4612 standard model

The second recommended PC is “THIRDWAVE Pro WORKSTATION X4612 standard model”.

OS	none
CPU	Intel Xeon Silver 4210R (rated 2.40GHz/3.20GHz/13.75MB/10Core/20Thread at TB) x2
memory	32GB (DDR4-3200 ECC RDIMM/16GB×2)
storage	No disc (2.5″ rear bay 1)
GPUs	NVIDIA T600 4GB (MiniDisplayPort x4) x 1 [Order]
price	Price starts at 72,8860 yen (as of 02/08/2022)

The THIRDWAVE Pro WORKSTATION X4612 standard model is a high-end model that achieves expandability and powerful performance. *Since there is no OS, you will have to choose by yourself.

Up to two NVIDIA® Quadro® and NVIDIA® GeForce® series ultra-high-end graphics cards can be installed.

In addition, assuming use on the desk side, it can be operated with a commercial 100V power supply, and can be used for various purposes such as high-resolution video/audio editing, deep learning , CAE/CAD, and 3D animation. .

③HP ZBook Studio 15.6inch G8 Mobile Workstation New standard model

The third recommended PC is “HP ZBook Studio 15.6inch G8 Mobile Workstation New Standard Model”.

OS	Windows 10 Pro (Japanese) (Downgrade from Windows 11 Pro)
CPU	Intel® Core™ i7-11800H processor (max frequency 4.6GHz, 8 cores/16 threads, 24MB cache)
memory	16GB DDR4-3200 (onboard)
storage	512GB M.2 SSD (PCIe, NVMe, SED OPAL2, TLC)
GPUs	Intel® UHD Graphics and NVIDIA T1200 (4 GB GDDR6)
display	15.6 inch wide full HD liquid crystal display (matte panel, maximum resolution 1920 x 1080, maximum brightness 400cd/m², maximum 16.77 million colors, IPS method, LED backlight, ambient light sensor)
price	Price starts from 35,2000 yen (as of 02/08/2022)

The next-generation Intel® Core™ i9 vPro® processor in the PC is designed to handle complex multi-threaded applications, making multitasking easy.

The HP ZBook Studio 15.6inch G8 Mobile Workstation New Standard Model is a laptop designed for performance workflows in every aspect, from keyboard to screen.

Also, up to NVIDIA RTX™ A5000 or GeForce RTX™ 3080 GPUs can be installed. So you can seamlessly render, design and multitask even with heavy files.

And with NVIDIA RTX™ professional graphics, the PC can query millions of rows of data sets and analyze them in real time, making it the perfect PC for data scientists and business intelligence professionals.

Summary

How was it?

This time, I explained what deep learning is and the difference between a PC for deep learning and a normal PC.

A PC has various parts and I think it is difficult, but I would like you to acquire knowledge by all means.

Follow us on Facebook for updates and exclusive content! Click here: Each Techy

June 10, 2024