Skip to content

SeongilHeo/CS6350

Repository files navigation

This is a machine learning library developed by Seongil Heo for CS5350/6350 in University of Utah.

Homework 1 - Decision Tree

Options

Option Description
--help Show this help message and exit.
--data DATA Choose Dataset: car, bank. (Default: car).
--dir DIR Directory of the data folder. (Default:./Data. Example: /path/to/Data).
-M, --miss Enable this flag to handle "unknown" values in the data.
-T, --tree Enable this flag to visualize the decision tree.
-D DEPTH, --depth DEPTH Set the max depth of the ID3 decision tree. (Default: 1 to Max).
-C CRITERION, --criterion CRITERION Set the criterion for impurity: information_gain, majority_error, gini_index. (Default: All criteria are used).

Homework 2 - Ensemble Learning

Options

Option Description
--help Show this help message and exit.
--dir DIR Directory of the data folder. (Default:../Data. Example: /path/to/Data).
--data DATA Choose Dataset: bank. (Default: bank).
-M, --model Choose ensemble model: AdaBoost, BaggedTrees, RandomForest. (Default: AdaBoost)
-N NUM, --num NUM Set number or range of # trees . (Default: 50, ex: 50, 1-50)
--ratio RATIO Set ratio of boostrapping. (Default: 1)
--seed SEED Set random seed for boostrapping. (Default: None)
--nattr NATTR Set bootstraping number of attribute for Randomforest. (Default: 2)
-q Q hoose question number. (options: 2a, 2b, 2c, 2d, 2e)

Homework 2 - Linear Regression

Options

Option Description
--help Show this help message and exit.
--dir DIR Directory of the data folder. (Default:./Data. Example: /path/to/Data).
--data DATA Choose Dataset: car, bank. (Default: car).
-M, --model hoose batch or stochastic Gradient Descent. (options: batch, stochastic)
-R, --r Set learning rate. (Optional, example: 0.01)
-q Choose question number. (options: 4a, 4b, 4c)

Homework 3 - Perceptron

Options

Option Description
--help Show this help message and exit.
--dir DIR Directory of the data folder. (Default:../Data. Example: /path/to/Data).
--data DATA Choose Dataset: banknote. (Default: banknote).
-M, --model Choose model. (Default: standard, Options: standard, voted, average)
-T EPOOCH,--t EPOOCH Set maximum epoch. (Default: 10)
-R Learningrate,--r Learningrate Set learning rate. (Default: 0.01)
-q Choose question number. (Options: 2a, 2b, 2c)

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published