Machine-Learning

it's a package of some basic machine learning algorithms,

I made it step by step,welcome to give me a hand

there are more information about the code on my blog

INTRODUCTION

· 1.linear_model.py

It contains three different kinds of linear models

They are Linear Regresiion、Logistic Regression and Linear Discriminant Analysis

There are more details in : detail document about the algorithms

the result of the py file:

· 2.decision_Tree.py

It's a Decision Tree algorithm's file,and actually, I am not so satisfied with the code

It's not easy to give the visuable result of the algorithm,if you could help me with the

code and visualable ,it's thankful

There are more details in : detail document about the algorithms

· 3.BP_neural_network.py

The code of the BP_neural_network.py is pretty good ,and I test it using the 'xor'

There are more details in : detail document about the algorithms

· 4.support_vector_machine.py

when i study the Support Vector Machine ，I think it a little hard to figure out the

model. so I use a package of the python to help me complete the code, And It's shoud be right

There are more details in : detail document about the algorithms

the result of the py file:

· 5.naive_bayes_classifier.py

I use a famous data set in the naive bayes classifier called 'pima-indians-diabetes dataset'

and the accurancy could be 70% when the average is 65% , I give the link to the set,you could try by yourself

There are more details in : detail document about the algorithms

· 6.AdaBoost.py

I use the same data set with the bayes algorithm in the AdaBoost，It shows a greater function than the bayes

and the accuracy could be 82%

There are more details in : detail document about the algorithms

· 7.K-means.py

The famous clustering algorithm ,K-means , It's easy to understand and easy to implement

I test the code by some dots creating by myself

There are more details in : detail document about the algorithms

the result of the py file:

· 8.pca.py

I combine the pca and K-means inthis file ,but it's even more diificult to test the fuction

than the decision tree , I make some resluts just for reference, if you have better way, please

help me ,thank you. and the data set I use is here

There are more details in : detail document about the algorithms

the result of the py file:

9.FastICA.py

I use FastICA to split the sound mixed with three different voices(baby/women/women), the two others are noise,firstly I make them the same size, and mix them with each other to create three new sounds,and then,using ICA to get the three original sound . It has a pretty good effect

you could fun much nore details from here : detail document about the algorithms

the mixing sound,and FastICA resault:

A10.SFA.py

SFA,slowness feature analysis,a very useful algorithm in feature engineering.I combine it with different classifiers to identify the humen face of different dataset,the yale_faces and orl_faces，you could choose the different dataset by filling '1' or '2'

you could fun much nore details from here : detail document about the algorithms

the yale's resault,and orl's resault:

CONTRIBUTOR

---

                                    CANTACT CARD
                           Author:     LiangjunFeng
                           Blog:       http://my.csdn.net/Liangjun_Feng
                           E-mail:     [email protected]
                           School:     zhejiang University
                           Begin from: 2017/8

---