Coursera-IBM-Machine-Learning-with-Python-Final-Project : The Best Classifier , Instructor - Saeed Aghabozorgi :
@ Task - loading a historical dataset from previous loan applications, cleaning the data, and applying different classification algorithm on the data.
Analyzing the neighborhood data and building the best classifier as prediction models using the different classifications algorithms - k-Nearest Neighbour(KNN) , Decision Tree , Support Vector Machine(SVM) , Logistic Regression.
The results is reported as the accuracy of each classifier, using the following metrics when these are applicable: Jaccard index , F1-score , LogLoss.
1.) Name - Building model using K-Nearest Neighbour (KNN)
Description - Building model using KNN Classifier, to predict whether a loan case will be paid off or not. Finding the best 'k' and accuracy evaluation, using the following metrics when these are applicable: Jaccard index, F1-score, Log Loss.
2.) Name - Building model using Decision Tree
Description - Building model using Decision Tree, to predict whether a loan case will be paid off or not. Finding the best 'k' and accuracy evaluation, using the following metrics when these are applicable: Jaccard index, F1-score, Log Loss.
3.) Name - Building model using Support Vector Machine (SVM)
Description - Building model using SVM, to predict whether a loan case will be paid off or not. Finding the best 'k' and accuracy evaluation, using the following metrics when these are applicable: Jaccard index, F1-score, Log Loss.
4.) Name - Building model using Logistic Regression
Description - Building model using Logistic Regression, to predict whether a loan case will be paid off or not. Finding the best 'k' and accuracy evaluation, using the following metrics when these are applicable: Jaccard index, F1-score, Log Loss.
Final Project Notebook - The Best Classifier