Wayfair-customer-data-analysis-competition
Motivation
To partake in an online competition of the famous home decor giant that involved data analysis and application of Classification and Regression models.
Goals
- To apply multiple Machine Learning models on real world data
- Data Wrangling, Visualizations and hyperparameter tuning.
- Presentation and analysis
Technical Achievements
- Advanced pre-processing
- Application of XGBoosting, LDA, QDA, SVM, Random forest, AdaBoosting, Ridge Regression and LASSO regression and preparing the data for each model.
- Comparison of all the models after 10 fold cross validated hyperparameter tuning using various evaluation metrics.
Challenges
- Deciding how to handle ordinal, categorical and missing data.
- Model Selection
Future Considerations
- Heavier Data Visualizations
- More advanced Regression evaluation techniques