Machine learning Notes (i)--Machine learning basics

1. What is machine learning

Machine learning is the conversion of unordered data into useful information.

The main task of machine learning is to classify and another task is to return.

Supervised learning: It is called supervised learning because such algorithms must know what to predict, that is, the categorical information of the target variable. (Classification, regression)

Unsupervised learning: Data does not have category information and is not given a target value. (clustering, density estimation)

Supervised Learning:

K- neighbor algorithm, linear regression, naive Bayesian algorithm, local weighted linear regression, support vector machine,Ridge regression, decision tree,Lasso minimum regression coefficient estimation

Unsupervised Learning:

K- means, maximum expectation algorithm,DBSCAN,parzcn window design

2. How to choose the right algorithm

Consider two questions: first, the purpose of using machine learning algorithms, what kind of tasks do you want the algorithm to accomplish, and what data to analyze or collect.

3. Steps to develop a machine learning application

(1) Collect data

(2) Prepare input data

(3) Analysis of input data

(4) Training algorithm

(5) Test algorithm

(6) using the algorithm

4. Python

First, install Python(installed in the Windows System)

: https://www.python.org/downloads/windows/

Select version:Python2.7.9, Win32 version

Ii. installation of NumPy and SciPy

: http://www.scipy.org/scipylib/download.html

Click on the Red box location, go to the download page, select the corresponding version to download and install (Numpy-1.9.2-win32-superpack-python2.7.exe, Scipy-0.15.1-win32-superpack-python2.7.exe)

Third, installation matplotlib

: http://matplotlib.org/downloads.html

Installed version: Matplotlib-1.4.3.win32-py2.7.exe

Machine learning Notes (i)--Machine learning basics