No Free Lunch (NFL) theorem
David Wolpert demonstrated that if you make absolutely no assumption about the data, then there is no reason to prefer one model over any other.
Instance-Based Versus Model-Based Learning
Supervised learning
需要人工标注数据
Unsupervised learning
不需要人工标注数据,能获取数据之间的潜在关系
semi-Supervised learning
一部分数据是需要人工处理的,一部分不需要
Active learning
Batch learning
训练数据分批
reinforce learning
Online learning
feature engineering
• Feature selection: selecting the useful features to train on existing features.
• Feature extraction: combining existing features to produce a more useful one.
• Creating new features by gathering new data.
machine learning challenge
1.不够充分的训练数据
2.没有表征的训练数据
3.质量差的训练数据
4.数据中不相关的特征对数据造成干扰。
5.模型训练的过拟合现象.