This is my final project for the IBM Machine Learning Course on Coursera. The Notebook and, within which, the dataset and the initial guidelines are provided by IBM for the course.
In this project, I used two pieces of a dataset. One of them is used to explore 4 different types of classification models and choosing the best version of each of them:
⋅ K Nearest Neighbors (KNN), in which I found the best K-value;
⋅ Decision Tree, in which I found the best maximum depth;
⋅ Support Vector Machine (SVM), in which I found the best kernelling function;
⋅ Logistic Regression, in which I found the best solver and inverse regulation value (C).
The other one is used to test the 4 models and compare their accuracy score using 4 different types of accuracy calculation methods:
⋅ Jaccard Similarity Score;
⋅ F1-Score;
⋅ LogLoss for the logistic regression.