Tuesday, May 28, 2024
HomeAIExplaining mathematics for AI | Three learning methods and learning steps are...

Explaining mathematics for AI | Three learning methods and learning steps are also introduced!

Table of Contents 

  • Mathematics is essential for learning AI!
    • Why we need mathematics in AI
    • there are exceptions
  • Three areas of mathematics closely related to AI learning
    • probability statistics
    • linear algebra
    • Calculus (analysis)
  • 3 steps to master the mathematics necessary for AI
    • 1. Basic mathematics for middle school and high school
    • 2. 1st and 2nd year college mathematics
    • 3. advanced mathematics
  • 3 ways to learn math
    • 1. learn from books
    • 2. learn online
    • 3. learn at school
  • Comparing different math learning methods
  • How to avoid frustration in learning mathematics
  • summary

Mathematics is essential for learning AI!

Studying mathematics is inevitable when studying machine learning. Here, I will explain why you need to know mathematics.

Why we need mathematics in AI

In conclusion, mathematics is necessary because understanding of “the mechanism of AI, that is, how AI learns, predicts, and evaluates” will progress at once.

The AI ​​mechanism uses knowledge from various fields of mathematical science, such as statistics, machine learning , and computer science. Mathematics is an indispensable common language in any field.

Therefore, it can be said that knowledge of mathematics is essential if you want to properly understand AI from the mechanism part.

there are exceptions

The programming language Python, which is often used for AI learning, has many useful libraries. By using these functions, you can create high-performance prediction models and image recognition models without understanding the contents of the library.

Therefore, it is not necessary to understand the mechanism of AI, and learning mathematics is not essential for those who only use AI as a tool.

However, knowledge of mathematics is essential when considering questions such as “Why did this result come about?” for AI learning results.

Three areas of mathematics closely related to AI learning

Here are three areas of mathematics that are relevant to understanding AI.

probability statistics

By using probability, concepts such as the distribution of data to be handled, the validity of predictions, and the accuracy can be mathematically expressed.

Statistics deals with various statistical information about a population. In addition, they can be used to refine the data and identify outliers in each data.

linear algebra

AI handles a large number of calculation formulas. And linear algebra is indispensable in order to express the calculation formula in an easy-to-understand manner.

Linear algebra makes it possible to easily express a large number of calculation formulas, making it possible to efficiently learn a large amount of knowledge.

Calculus (analysis)

Calculus is commonly used to express the “slope” of a function.

In the AI ​​field, “slope” refers to “error”. “Error” is the difference between the correct value and the predicted value when making a prediction.

In order to aim for more accurate AI, it is necessary to bring this “error” closer to 0. To calculate the “error”, calculus is essential.

3 steps to master the mathematics necessary for AI

So what are the steps in learning mathematics? This chapter introduces the fields and steps of mathematics required for AI learning.

1. Basic mathematics for middle school and high school

It is not necessary for everyone to re-study from the basic mathematics of junior high school and high school.

Certainly, basic knowledge of mathematics is necessary to understand mathematics at the university 1st and 2nd year level. However, most online courses, books, etc., which will be described later, start with content like a review of junior high and high school mathematics.

For those who are not confident in mathematics or who stumbled while studying, we recommend reviewing from junior high school and high school textbooks.

2. 1st and 2nd year college mathematics

At a science university, first and second year students study linear algebra and calculus as general education subjects .

By mastering the basics of these liberal arts subjects and mastering the basics of probability and statistics in addition to them, it is possible to understand the basic mechanism of AI.

In other words, these fields can be said to be the main mathematics necessary for AI.

3. advanced mathematics

By learning more advanced mathematics, it becomes possible to have a deeper understanding. Examples include Bayesian estimation, gradient descent, and Lagrange’s method of undetermined multipliers.

In addition, even if we say AI in a nutshell, there are various fields within AI ( natural language processing , image recognition , etc.). The knowledge you need depends on which field you are studying.

We recommend that you learn the necessary mathematics each time.

3 ways to learn math

Here are some of the best ways to study mathematics. Find a study method that works for you and stick to it.

1. learn from books

The first method is to learn from books.

This method is recommended for those who can read the book on their own. Books are also rich in information.

If you want to learn mathematics more deeply and carefully, or if you want to learn the essence, learning from books may be suitable.

Below are four recommended books.

1. Mathematics book for artificial intelligence programming

 

This is a book where you can learn the mathematics necessary for AI from the basics. Recommended as an introductory book for those who are resistant to mathematics.

2. Mathematics of Deep Learning in the Shortest Course

This is also a great book to learn from the basics.

Also, since the code written from scratch using mathematics is provided in “Jupyter Notebook format”, you can learn while actually moving.

It is also recommended for those who are interested in implementation.

3. Deep Learning from Scratch ―Theory and Implementation of Deep Learning in Python

Difficulty☆☆

The knowledge of AI mathematics, which is more complicated than the above two books, is explained simply. Recommended for those who understand basic mathematics.

Four. Machine Learning with Bayesian Inference (Introduction)

 

This is a book where you can deeply learn the probability statistics necessary for AI.

Recommended for those who understand the above-mentioned mathematics books and want to study more.

2. learn online

The second method is to study online.

This method is recommended for those who find it difficult to learn from books alone and want to use their time efficiently.

There are also online courses where you can take lectures by famous instructors and ask questions. Here are three of our favorite online courses.

1. Udemy

Udemy

Udemy is an online learning platform with over 100,000 courses. Once you purchase a course, you can study at your own pace with no study deadlines.

Here are some of my favorite Udemy courses:

[Kikagaku Style] Artificial Intelligence/Machine Learning Leaving the Black Box Course – Beginners –[Kikagaku Style] Artificial Intelligence/Machine Learning Leaving the Black Box Course – Intermediate Edition –Mathematics Course for AI: Linear Algebra/Probability/Statistics/Differentiation for Artificial Intelligence

In this course, you can learn mathematics along with AI knowledge and terminology. You will be able to understand the principles of machine learning from mathematics and implement them in programming ( Python ).

Mathematics course for AI: Learn slowly and carefully linear algebra/probability/statistics/differentiation for artificial intelligence . is clearly explained through

Besides, Udemy has a lot of courses, so you should be able to find a course that is perfect for you.

2. Chainer

Deep Learning Primer: Chainer Tutorial

A free online learning service that teaches everything from the basics of mathematics to deep learning.

Chainer has

  • differential
  • linear algebra
  • Probability/Statistics
  • Simple regression analysis and multiple regression analysis

There are university mathematics courses such as this, so you can learn the minimum necessary mathematics.

The recommended point is that it is designed to solve the problems that beginners tend to stumble on, ‘What should I learn first?’ and ‘What should I learn next?’

*The development of Chainer has ended, but the content of the mathematics course itself is easy to understand and polite, so it is recommended.

3. AI academy

From the beginning of programming to mathematics and statistics necessary for machine learning, and deep learning,

You can learn a wide range of knowledge from over 300 original texts.

Also, instead of just reading the text, you can proceed with learning in a quiz format, AI finds problems you are not good at, and you can efficiently overcome your weak areas.

After completing the mathematics course necessary for machine learning, you will be at a level where you can read formulas such as books related to machine learning.

3. learn at school

The third method is to learn at school.

This method is recommended for those who have high hurdles to self-study, want to create a learning pace, and want to learn programming consistently.

The following three schools are recommended.

1. AI job boyfriend

AI job boyfriend

AI Job Kare is a school specializing in AI technology where you can systematically learn from experienced active AI engineers and data scientists.

The school has many introductory courses, so even beginners can learn with confidence. There is also support unique to the school, so it is possible to learn without dropping out on the way.

The following courses are recommended for learning mathematics.

Python + Mathematics Super Introductory CourseMathematics courses necessary for Python + machine learning

You can choose a course according to your mathematics knowledge level.

AI job boyfriend

2. Skill up AI

Skill-up AI | Support for building AI human resource development and development organizations

Skill Up AI is a learning community where you can learn ultra-practical machine learning skills that can be used immediately after completing the course from data science specialists who are active at the forefront.

All courses can be taken online. You can actively ask questions according to the progress of the lecture in the same chat group as the face-to-face students,

Stay motivated and keep learning.

Among them, by taking the ” Basic Mathematics Course for Machine Learning/Deep Learning ” and ” Applied Mathematics Course for Machine Learning/Deep Learning “, even those who have no prior knowledge of mathematics can understand machine learning. You can grow to a directly connected level.

Since a free trial version is available, we recommend that you try it first to see if it suits you.

Comparing different math learning methods

  books Online course school
cost -¥ 10,000 Free – tens of thousands of yen -¥100,000
period weeks-months 1 week – several weeks 1 week – several weeks
difficulty ☆☆☆☆ ☆☆ ☆☆
Degree of recommendation ☆☆ ☆☆☆☆ ☆☆

How to avoid frustration in learning mathematics

In order not to be frustrated in learning mathematics, the most important thing is to choose a learning method that suits you, make an appropriate plan that is reasonable, and steadily proceed steadily.

Look at your goal, compare it with your current math level, and plan the necessary process!

Summary

This time, I have told you the necessity of mathematics, the reason for it, the necessary knowledge, and recommended study methods.

I hope that this article has helped you understand a little more about mathematics in AI learning and find a study method that suits you.

RELATED ARTICLES

What is an AI algorithm?

LEAVE A REPLY

Please enter your comment!
Please enter your name here

Recent Posts

Most Popular

Recent Comments