Do you want to become an AI practitioner? Systematic learning of machine learning is key!
-
What is Machine Learning?
Machine learning is a science that enables computers to function without explicit programming. Over the past decade, machine learning has provided us with practical tools such as self-driving cars, real-time speech recognition, and efficient web searches, greatly enhancing our understanding of the human genome. Many researchers believe that developing machine learning is the best way to advance towards human-level artificial intelligence.
Here are three steps to systematically learn machine learning:
-
Phase One: Basic Introduction
The first step is to study some classic courses along with classic books. Generally, this process is suitable to be completed within six months. The courses and books I introduce in this section are of very low difficulty, requiring little in terms of mathematics and programming.
Recommended Book One: Python Machine Learning
Recommended Book Two: Introduction to Statistical Learning with R (ISL)
Recommended Book Three: Zhihua Zhou’s Machine Learning
-
Phase Two: Advanced Learning
In this phase, you already have a basic understanding of machine learning. If you have seriously read ISL and completed Andrew Ng’s course, I believe you have theoretically understood linear regression, data compression, feature engineering, and have a theoretical foundation for simple regression/prediction problems. At this point, the most important thing is to practice!
1. Kaggle Competitions/Practice
Kaggle has many excellent datasets and competitions. You can try these challenges to achieve rankings and even win prizes, which will be very helpful for future job searches. Moreover, another great advantage of Kaggle is that users share their experiences and opinions, and you can also ask questions for others to help you with some corrections.
Recommended Book: 
2. Learning from Sklearn Documentation
The Sklearn documentation is one of the few technical documents that reads like a tutorial and is very valuable. For example, if you want to learn how to use logistic regression in Python, you can refer to the link:
Reference Link:: Logistic Regression 3-class Classifier
(Optional): Deep Learning (3-6 months)
Deep learning is currently a hot topic, and many companies are looking for talents in deep learning. Although deep learning is just a subset of machine learning, friends interested in this direction can study deep learning separately after completing the above learning.
-
Phase Three: Practical Experience
Congratulations! If you have completed the above plan, it means you have considerable machine learning capabilities. At this stage, the focus should be on forming a systematic knowledge framework. Remember, do not bite off more than you can chew!
Machine learning is not just about theory; the best way is to practice. Therefore, you should first try to engage in research, and then attempt internships in the industry. When it comes to research opportunities, take them if available; if not, it is not a great regret.
↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓↓
Subsequent Knowledge and Skills:
1. Basic linear algebra. If you haven’t learned it yet, it’s better to take this course before diving into research; otherwise, you might regret it. 2. Learning Python is sufficient. R can also be used. 3. English. At least be able to listen and read basic content, as there isn’t enough Chinese material, and it’s hard to avoid reading a lot of English materials. It’s recommended to look at the English version of tutorials when studying certain topics; YouTube can have subtitles.