In this Project Customer data is used to predict a label as a default or not.
The features are removed of nulls and visualized , then the data is changed to convert categorical variables to binary
variables using a one hot code technique. The data is then normalized and then the following models are ran :
K nearest neigbor, Decision Tree, Support Vector Machine, Logistic Regression. Each models accuracy is then visualized
using a Confusion matrix which shows the errors given by the model as False Positives and False Negatives. The Jaccard
index and F1 score is then plotted.