(1) Know some background knowledge of artificial intelligence;

(2) Supplementing knowledge of mathematics or programming;

(3) Familiar with the machine learning tool library;

(4) systematically learning AI knowledge;

(5) Do some AI applications;

1 Understand the background knowledge of artificial intelligence

There are many concepts in artificial intelligence, such as machine learning, deep learning, neural network and so on, which make beginners feel that artificial intelligence is mysterious and difficult to understand. At the beginning, it is enough to know the general meaning of these nouns, so you don't have to go too far. After studying for a while, I naturally know what these concepts represent.

Artificial intelligence is an interdisciplinary subject, in which mathematics and computer programming are the two most important aspects of learning artificial intelligence. These articles "Understanding Artificial Intelligence" before "Zhiyun AI Column" have also been introduced for everyone. Students who haven't seen it can go and have a look.

The following figure is the general route of artificial intelligence learning:

2 Supplementary knowledge of mathematics or programming

For engineers who have graduated, before systematically learning AI, they generally need to add some knowledge about mathematics or programming. If you are good at math and programming, it will be much easier to learn artificial intelligence.

Many students are afraid when it comes to mathematics, but learning artificial intelligence can say that mathematics can't be bypassed. There is no need for advanced mathematics in the introductory stage, mainly advanced mathematics, linear algebra and probability theory, which means that the mathematics knowledge of freshmen and sophomores is completely sufficient. If you want to work as a machine learning engineer or engage in the research of artificial intelligence, you should learn more about mathematics. Being good at math will be a great advantage in your job.

Python is very popular in the field of machine learning, which can be said to be the most used programming language, so Python programming also needs to be mastered. Among many programming languages, Python is relatively easy to learn and use, and learning Python well will also benefit a lot.

3 Familiar with machine learning tool library

Now people realize artificial intelligence, mainly based on some tool libraries of machine learning, such as TensorFlow, PyTorch and so on.

I recommend you to study PyTorch here. PyTorch is very popular and is an easy-to-use machine learning tool library. Some people say that PyTorch is "not very good, but it is very comfortable to use".

When you start learning artificial intelligence, you can run an example of an official website, such as MNIST handwriting recognition. This will have a perceptual understanding of artificial intelligence and eliminate the initial strangeness. Then you can look at the code inside, and you will find that the program of neural network is not complicated, but there are many problems about the principle and training of neural network. This is a good thing, because learning with questions will be more effective.

4 system learning artificial intelligence

Artificial intelligence here mainly refers to machine learning, because at present, artificial intelligence is mainly realized through machine learning.

Machine learning knowledge mainly has three parts:

(1) Traditional machine learning algorithms, such as decision tree, random forest and SVM, are called traditional machine learning algorithms, which are relative to deep learning.

(2) Deep learning refers to deep neural network, which can be said to be the most important and core artificial intelligence knowledge at present.

(3) Reinforcement learning, which originated from cybernetics, is sometimes translated into reinforcement learning. Deep learning can be combined with reinforcement learning to form deep reinforcement learning.

What needs to be known here is that deep learning is not difficult to learn. For some engineering graduate students, it usually takes only a few weeks to get started and train some practical neural networks. However, it is not easy to have a deep understanding of deep learning, which usually takes several months.

There are many traditional machine learning algorithms, some of which have many mathematical formulas, such as SVM. These algorithms are not easy to learn. You can learn deep learning first, and then slowly supplement these traditional algorithms.

Intensive learning is more difficult, and it usually takes two or three months to understand.

5 do some artificial intelligence applications.

After several weeks of deep study, you can try to do some AI applications, such as image recognition, style transfer, text poetry generation and so on. Learn a lot better while practicing, and you will gradually deepen your understanding of neural networks.