Stanford Machine Learning Course Note (1) Supervised learning and unsupervised learning

The last three weeks of Andrew Ng's machine learning were recently followed by the linear regression (Linear Regression) and logistic regression (logistic Regression) models in machines learning. Make a note here.
Also recommended a statistical study of the book, "Statistical Learning method" Hangyuan Li, Book short, only 200 pages, but the content is basically covered the theoretical basis of machine learning.

Notes <1> Learn more about supervised learning and unsupervised learning

Machine learning : It is a subject of computer -based probabilistic statistical model of data construction and the use of models to predict and analyze data .

machine Learning Algorithms Classification : Supervised learning (supervised learning), unsupervised learning (unsupervised learning), semi-supervised learning (semi-supervised learing) and intensive learning (reinforcement learning).

Supervised learning

• Supervised learning learns the model from the training data collection and predicts the test data.
• The purpose of supervised learning is to learn a mapping from input to output, that is, to find such a model (also known as hypothesis hypothesis) that belongs to the set of mappings from the input space to the output space.
• When the input and output variables are continuous variables, the prediction problem is called regression (regression) problem; The prediction problem of the output variable is a finite discrete variable, which is called the Classification (classification) problem; The prediction problem of both the input and output variables is a sequence of variables called callouts ( tagging) problem.

(The Housing price prediction and Breast cancer examples in the course describe regression and classification issues respectively)

Unsupervised learning

• Unsupervised learning (unsupervised-learning) is an algorithm (algorithm) for artificial intelligence networks, which is designed to classify the original data in order to understand the internal structure of the data. Unlike supervised learning networks, unsupervised learning networks are not aware of the correctness of their classification results when they are learning, that is, not supervised enhancement (telling it what kind of learning is correct). The feature is that only the input paradigm is provided for this network, and it automatically identifies its potential class rules from those examples. When the study is complete and tested, it can also be applied to new cases.

• A typical example of unsupervised learning is clustering. The purpose of clustering is to bring together things that are similar, and we do not care what this class is. Therefore, a clustering algorithm usually needs to know how to calculate the similarity to begin to work.

Semi-supervised learning

Intensive Learning

Copyright NOTICE: This article for Bo Master original article, without Bo Master permission not reproduced.

Stanford Machine Learning Course Note (1) Supervised learning and unsupervised learning