人类认识世界获取知识有多种途径，类似地，机器学习也有多种途径，这些方法

There are many ways for humans to understand the world to acquire knowledge. Similarly, machine learning also has many ways. Which of these methods is better or worse depends on the application environment, which is a matter of opinion [3]. After studying supervised learning, decision trees, and multi-layer perceptrons, students often ask: Which method is the best for classification? Which method is the most efficient? Why is the training error of my algorithm very small but the effect is not ideal? Facing similar questions raised by students, teachers often encourage students to find more literature, do experiments, deepen their understanding of knowledge from experiments, deepen their understanding of the nature of the problem from practice, and possibly discover new problems. Encourage students to read the literature with questions and hypotheses to do experiments, and share their own reading experience and experiments with everyone. The communication methods include classrooms, machine learning seminars, WeChat groups, Tencent conferences, etc.

There are many ways for humans to know the world to acquire knowledge, and similarly, machine learning has many ways, and the differencebetween between these methods depends on the application environment, which is a matter of insight. After learning supervised learning, decision trees, and multi-layer sensors, students often ask: Which method works best for classification? Which method is most efficient? Why is my algorithm having little training error and not having ideal use? In the face of similar questions raised by students, teachers encourage students to find more literature, do experiments, deepen understanding of knowledge from experiments, deepen understanding of the nature of problems from practice, and may find new problems. Encourage students to carry out experiments with questions and assumptions, and share their own literature and experiments with you in a variety of ways, such as classroom, machine learning seminars, WeChat group, Tencent conferences, etc.

There are many ways for human beings to understand the world and acquire knowledge. Similarly, there are also many ways for machine learning. Which of these methods is better or worse depends on the application environment. It is a matter of different opinions [3]. After learning supervised learning, decision tree and multi-layer perceptron, students often ask: which method is the best for classification? Which method is the most efficient? Why my algorithm training error is very small, but the use effect is not ideal? In the face of similar problems raised by students, teachers encourage students to search more literature, do experiments, deepen the understanding of knowledge from the experiment, deepen the understanding of the nature of the problem from practice, and may find new problems. Students are encouraged to read literature with questions and hypotheses, and share their reading experience and experiment with you. The communication methods include classroom, machine learning workshop, wechat group, Tencent conference and so on