Fake-Instagram-Profile-Detection

This is a mini-project I completed under Coursera Project Network.

In this hands-on project, I have built a deep neural network model to detect spam (fake) Instagram accounts.

• Dataset Used - Uploaded as insta_train.csv and instra_test.csv

  Parameters Used

  Feature	Values	Type of value
  Profile Pic	Yes or No	0 or 1
  Full Name	No. of words	Integer
  Description Length	No. of words	Integer
  Private	Yes or No	0 or 1
  Posts	No. of posts	Integer
  Followers	No. of followers	Integer

• Exploratory Data Analysis - Checked for null values and analysed the statistical summary of the data.
• Data Visualisation - Using Seaborn and Matplotlib libraries in Python
• Data Preprocessing - Data Normalisation using Scikit Learn and transformed 'Fake' column data to categorical using Keras
• Creating model - Sequential Model. Used ReLU activation function for the input and the hidden layers and Softmax activation function for the output layer.
• Training the model - Used 20 epochs for training the model.
• Evaluation - Getting an accuracy of 91%.