For my classification project I chose to create a model that is able to predict credit card fraud based on 29 datapoints that were collected and provided in a Kaggle dataset. Obviously in the case of credit card fraud, I wanted to get as high of accuracy as possible without overfitting the model and it becoming useless when put out into the real world. So that is what I set out to achieve, a model that is able to predict credit card fraud with the highest degree of accuracy.
My final model shows the accuracy I was able to achieve. The results are from a population that was completely different from that which was used to train and test the model. I elected to not clean the data for this test to see how the model would handle any sort of unexpected values, although the dataframe was very clean to begin with.