edu.illinois.cs.cs125.fall2019.mp

Class WekaTrainer

java.lang.Object

edu.illinois.cs.cs125.fall2019.mp.WekaTrainer

public class WekaTrainer
extends java.lang.Object

Workbench for training the Weka machine learning model that distinguishes random walks from human paths.

This is only needed in Checkpoint 5.

CAUTION: trainingpaths.json will not exist in the app or during official grading.

Constructor Summary

Constructors

Constructor and Description
WekaTrainer()

Method Summary

Modifier and Type	Method and Description
static void	main(java.lang.String[] args) Entry point to allow running this class outside Android.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

WekaTrainer

public WekaTrainer()

Method Detail

main

public static void main(java.lang.String[] args)
                 throws java.lang.Exception

Entry point to allow running this class outside Android.

CAUTION: When running outside of Android or Robolectric, it is not possible to use Android classes or any methods that rely on them. When running inside Android, it is not possible to access the development computer's filesystem. This method should only be used to train the model which gets baked into the app; it must not be called from app code.

Parameters:

args - command-line arguments (unused)

Throws:

java.io.IOException - if something goes wrong loading the JSON file or saving the model

java.lang.Exception - if something goes wrong serializing the classifier