SurrajKumar2000 / Fake-Profile-Detection

This repository is for detection of fake profiles in online social networks by the use of ML prediction techniques such as Random Forest, Neural Network and Support Vector Machine(SVM).

Fake-Profile-Detection

A model to detect fake user profiles using only the visible data as mentioned. This is a project where training models

We have trained them in SVM, RandomForest and a 3-layered ANN which gave an accurary of 93%-98% on the test set.

REQUIREMENTS:

        matplotlib==1.4.3
        pandas==0.16.2
        pybrain==0.31
        numpy==1.9.2
        ipython
        sexmachine==0.1.1
        pip==6.0.8
        scikit-learn==0.17

CONCLUSION:

Among all other classifier methods the RandomForest classifier method gives greater accuracy.