Skip to content

ghazalb76/Classification-SVM

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

5 Commits
 
 
 
 
 
 

Repository files navigation

Classification-SVM

In this project, I have used Insurance Claim Dataset, related to the transactions of an insurance company in the United States to detect fraud. Following steps show the implementation process in detail:

  • Preprocessing: One-Hot Encoding and filling or removing missing data
  • Seperating train data and test data
  • Generating primary SVM
  • Evaluating the model using Confusion Matrix for each of the 4 kernels: linear, poly, rbf, sigmoid
  • Optimizing values of the parameters using "GridSearchCV()" method
  • Regenerating the SVM based on best optimized parameters
  • Drawing ROC for both primary and final SVM model

In addition, I have degraded the features' degree to 2D using PCA:

By running the code using following command, you can see the results by yourself:

    python src.py

But before that you need to have the following Python packages installed:

  • pandas >= 0.25.1
  • numpy >= 1.17.2

A very complete Persian Report is also included in Report.pdf.

Thanks to Mr. Josh Starmer for his wonderful tutorials :D

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages