Skip to content

A descriptive comparison between different supervised classifier and neural network

Notifications You must be signed in to change notification settings

ritvikkhanna09/Census-classifier-comparison

Repository files navigation

Classifier Comparison on the Census data

This project is a descriptive analysis of various supervised learning classifiers like Decision Tree, Random Forest, Naive Bayesian, Support Vector Machines and Neural Network.

Dataset

  • The dataset is taken from the UCI repository from here.
  • The dataset contains 14 attributes describing an individual personal and working condition.
  • The class label is income column giving the information whether an individual's yearly income is less than or greater that equal to 50K.

Getting Started

  • The code is written in python using various libraries like pandas, numpy and sklearn.
  • I have used Jupyter notebook for easier understanding and execution of the codes.
  • Tutorial on how to use Jupyter notebook

If any libraries are missing on the system simply pip install them and re-run the code

pip install <module>

Project Files

I have divided the code into 6 jupyter files where the first 5 files are running the a specific classifier and the compiler file simply compares the results and plot the data.

The files corresponding to each classifier is as following:-

Results

  • Confusion matrix

    alt text
  • ROC curve

    alt text
  • Scores

    alt text

Author

Ritvik Khanna

contact

Releases

No releases published

Packages

No packages published