Skip to content

sudeepgbits/machine-learning

 
 

Repository files navigation

machine-learning

CS 7641 coursework

Data Sources

Prerequisites:

  • Install all the dependent python libraries - scikit-learn, matplotlib, numpy and pandas
  • util.py provides utlity function to input data and print learning curves abstracted for other files
  • Store the .py files and data in the same folder

Decision Tree:

  • In DecisionTree.py, remove line 85 and replace it with a function call to draw_learning_curve_1() to draw_learning_curve_2() to create the learning curves for Phishing Dataset and Optical Recognition Dataset, respectively. Modify line 84 with X1, Y1 or X2, Y2 to pick the dataset.
  • Additionally, a 3-D graph can be generated with the function testBothParams() and max depth graph can be generated by calling testMaxDepth()

KNN:

  • Follow same steps as Decision Trees by calling functions to create graphs using KNN.py

SVM:

  • Use either plot function to plot the graphs

Neural Network : Multilayer Perceptron:

  • The code in MultiLayerPerceptron.py generates epoch curves for Phishing Data, uncomment the last block and comment lines 85 to 98 to generate them for Optical Recognition Dataset

Boosting:

  • Uncomment line 41 to 62 to create graphs for Optical Recognition Dataset, change learning_rate or max_depth to create graphs for a higher or lower learning_rate or depth
  • Comment the block for Phishing or create two sub plots

References:

About

CS 7641 coursework

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%