Machine Learning



Features

  • learn from a training set to classify/categorize data examples
  • predict the class/category of unfamiliar (novel) data examples based on what is learned
  • evaluate the training accuracy and effectiveness of different machine learning models on your training set
  • example applications:
    • train & predict membrane transport based on pharmacological profiles
    • train & predict disease of patients based on physiological parameters
    • train & predict behaviour based on physiological parameters & preferences
    • etc...
  • Author: Mingyuan Li


System Requirements

Operating System: Windows, Mac OS, Linux, or Solaris
Disk Space Usage: 7 MB
RAM: depends on your training/test/data set size
Java: Java Runtime Environment Version 8u45 or above


Screenshots

Click on an image below to see the image in its full size.

Shortcut: you can use the left and right arrow keys to traverse through the screenshots :)


Machine learning classifiers screenshot Machine learning classifiers screenshot Machine learning classifiers screenshot Machine learning classifiers screenshot

Specifications

Training / test / data set Formatting requirement Example download
Training set
  • each data attribute must be separated by commas "," (use .csv file type)
  • each input example must begin from a new line
  • class/category of data: the value in the first column (or the first value in each line); this can contain alphanumeric characters
  • training data: (starting from second column) must be only numbers. (empty values are currently treated as 0, but this can be changed)
  • hint: you can use numbers to replace words (ie. denote 1 as male and 0 as female, etc.)

 View sample training data
Download sample training data
Test set
  • used to test the accuracy/effectiveness of each machine learning model on your training set
  • the test set needs to be formatted the exact same way as training set
  • hint: you can split your training set into training set and test set

 View sample test data
Download sample test data
Data set
  • this data set contains only data (numbers only), without classes/categories
  • the machine learning program will classify/categorize your data based on what is learned from training set
  • hint: this data set should have 1 less column than your training/test set

 View sample data
Download sample data

click here to view output for the sample inputs


Information on example data

  • the sample data provided trains the program to determine what a person might do on a particular day (soccer, ski, or stay in doors) based on the day's temperature, wind speed, and precipitation.
  • the 1st column in training and test set is what a person would do on a particular day
  • the 1st data attribute (2nd column in training and test set; OR 1st column in data set) is the temperature (°C) of a particular day
  • the 2nd data attribute (3rd column in training and test set; OR 2nd column in data set) is the wind speed (km/hr) of a particular day
  • the 3rd data attribute (4th column in training and test set; OR 3rd column in data set) is whether it is raining/snowing or not
    (0 = no precipitation, 1 = raining, 2 = snowing)
  • all files are of .csv format
  • note: data of the same class/category do not have to be grouped together in training and test sets



Download

This software is no longer available for download. Please email me at mingyuanli.5000[(at)]gmail.com if you wish to have a copy of the software.