Data-Science-Projects

This repo is dedicated to independent learning through various data science projects

Project 1: Exploring KNN with Classifying Fradulent Credit-Card data

This project analyzes credit card data (from kaggle) using K-Nearest Neighbors. The goal is to correctly classify fradulent data from verified transactions. Below shows the skills highlighted in the Exploring KNN with Credit Card Data juptyer notebook:

• SKlearn Implementation
• Homegrown Implementation
• Pipelines
• Hyperparameter tuning

Libraries

• NumPy
• Pandas
• Sklearn

Project 2: HCDR Project (Home Credit Default Project)

This project is based on the Home Credit Default Risk (HCDR) Kaggle Competition. The goal of this project is to predict whether or not a client will repay a loan. In order to make sure that people who struggle to get loans due to insufficient or non-existent credit histories have a positive loan experience, Home Credit makes use of a variety of alternative data--including telco and transactional information--to predict their clients' repayment abilities.

This project utilizes important data science concepts. Below highlights the skills required:

1. EDA (through pandas dataframes and graphics)
2. Feature Engineering (including numeric and text data)
3. Establishing pipelines
4. Feature Selection (including k-best and decision trees)
5. Model Selection (hyper-parameter tuning the following)
   • Logistic Regression
   • K-Nearest Neighbor
   • Support Vector Machine
   • Stochastic GD
   • Random Forest
   • Light GBM
   • XGBoost
   • Multi-layer Perceptron

The most successful model was MLP that achieved a public and private kaggle score of 76%.