Requirements to run the source are:
- Python3 should be installed - if not visit https://www.python.org/downloads/
- Libraries required include: a. Pandas - for dataset manipulation - https://pandas.pydata.org/pandas-docs/stable/install.html b. sklearn - for machine learning algorithms - http://scikit-learn.org/stable/install.html c. matplotlib - for plotting result graphs - https://matplotlib.org/users/installing.html
For executing source code:
- Download dataset from https://www.kaggle.com/mlg-ulb/creditcardfraud and unzip it.
- In the folder where creditcardfraud.csv is present, make a folder with name source.
- Copy all source file in source folder.
- Each source file can run independently. For executing use any IDE, or for command line use command python3 filename.py
- For obtaining different models in same domain change parameters of model initialization in source files as given. E.g. Change model = svm.SVC(kernel='rbf', C=1, gamma=10) to model = svm.SVC(kernel='linear', C=1) to change SVM model from RBF to Linear.
The source files can take anywhere between 3s to 15min to execute for given dataset.