In this notebook I train and tune models to predict credit card fraud using the dataset from Kaggle (https://www.kaggle.com/dalpozz/creditcardfraud).
The dataset contains credit card transactions recorded over two days, about 280000 transactions. The transactions have been labelled as fraud or non-fraud. The other variables are time since the first transaction in the data set, amount of the transaction, and 28 additional variables that are the result of principal component analysis on unspecified predictor variables. This is in order to anonymise the data.