Machine Learning Algorithms

Several Machine Learning Algorithms

How To Setup

Netbeans

File -> New Project Select The Maven Folder Select Java Application Next, Name your project, Finish

Other

Create A New Maven Project (look it up if you have to)

Add Dependancies

Paste this inside the <project> tag of the pom.xml file:

<dependencies>
    <dependency>
        <groupId>com.googlecode.matrix-toolkits-java</groupId>
        <artifactId>mtj</artifactId>
        <version>1.0.2</version>
    </dependency>
</dependencies>

If you want to utilize JMatIo for reading .mat files, look up tutorials for adding local jars to Maven projects for your IDE. The JMatIO jar is in the /src/main/resources/JMatIO_071005/lib folder of the project. This link was helpful for the NetBeans IDE, but a quick google search will reveal results for other IDEs.

Add Files

Initialize a new git repository in the root directory of the maven project and pull all the files from this repository

—or—

Download all the files and manually paste them into the src/main/java directory of your project

Using Fmincg

Look at the MTJTests.java file for a demo. Basically, you define a cost and a gradient function in a new CostGradient Object and run the routine of the Fmincg object. Call .getX() from the output object to get the optimal values.