Feature engineering and forecasting modeling for the Kaggle competition.
Algorithms used:
- Gradient Boosting
- LSTM Recurrent Neural Network
- Random Forest
- Extra Trees Regression
- Ridge Regression.
Model blending is used to acheive the best accuracy.
Solution ranked top 8% on Kaggle Competition (123rd out of 1675) on now-defunct account (Ortan Farnafe).