Skip to content

Fraud detection model implemnting weak learners and a simplified adaptive boosting algorithm. Classifies test data at 76% accuracy

Notifications You must be signed in to change notification settings

agene0001/Fraud-Detection

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

1 Commit
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Programming Assignment 7: Fraud Detection

/* *****************************************************************************
 *  Describe how you implemented the Clustering constructor
 **************************************************************************** */


/* *****************************************************************************
 *  Describe how you implemented the WeakLearner constructor
 **************************************************************************** */


/* *****************************************************************************
 *  Consider the large_training.txt and large_test.txt datasets.
 *  Run the boosting algorithm with different values of k and T (iterations),
 *  and calculate the test data set accuracy and plot them below.
 *
 *  (Note: if you implemented the constructor of WeakLearner in O(kn^2) time
 *  you should use the small_training.txt and small_test.txt datasets instead,
 *  otherwise this will take too long)
 **************************************************************************** */

      k          T         test accuracy       time (seconds)
   --------------------------------------------------------------------------

/* *****************************************************************************
 *  Find the values of k and T that maximize the test data set accuracy,
 *  while running under 10 second. Write them down (as well as the accuracy)
 *  and explain:
 *   1. Your strategy to find the optimal k, T.
 *   2. Why a small value of T leads to low test accuracy.
 *   3. Why a k that is too small or too big leads to low test accuracy.
 **************************************************************************** */


/* *****************************************************************************
 *  Known bugs / limitations.
 **************************************************************************** */


/* *****************************************************************************
 *  Describe any serious problems you encountered.
 **************************************************************************** */


/* *****************************************************************************
 *  List any other comments here. Feel free to provide any feedback
 *  on how much you learned from doing the assignment, and whether
 *  you enjoyed doing it.
 **************************************************************************** */

About

Fraud detection model implemnting weak learners and a simplified adaptive boosting algorithm. Classifies test data at 76% accuracy

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages