Machine_Learning

I learned some of the basic methods of machine learning, implemented all the procedures by hand-coding each method, and tested the model with simple data sets. Collect all the code here and the project will be continuously updated. Currently, matlab is temporarily used to complete the code.

The following algorithm has been implemented so far:

• Logistic Regression
  • SGD Logistic Regression
  • Newton Method Logistic Regression
• Least squares regression
• Naive Bayesian
  • Bernoulli Spam Classification
  • Multinomial Spam Classification
• Gaussian Discriminant Analysis
• Support Vector Machine(Use SMO algorithm)
• Back Propagation Neural Network

Dependencies

• Matlab

Data

The data needed for each project is placed in the project root directory. Of course, I used the same data set for many projects to compare the model. For this particular dataset, I put it in a separate dataset directory.

Training

For each project, run its corresponding .m file.

Result

All the models I tested worked well on the given data set.