In this project we will see how machine learning through text classification can be used to detect fraud in mobile money transfer platforms using the following tools:
- Python, Scikit-learn, Pandas, jupyter notebook
We are going to build an intelligent model which can be able to learn and tell in the future if a message is fraudulent or not, based on dataset from Kaggle.
The main parts of this project are:
-
1-Data Cleaning: clean data based on data analysis and data visualization
-
2-Data Preprocessing: preprocessing data (feature encoding,scaling,dimension reduction)
-
3-Model Building: building two models Random forest and SVM by training the model, tuning the parametrs and test
NB: data repository contains all the data, but it is not availabe in gitlab because of its huge size (1.6 GB)