Abalone Age Prediction - Random Forest
1. Random Forest was the winning model 2. 0.25 data split was the best split 3. Reducing the redundant 'sex' feature improved score 4. The best model state in this notebook gave a leaderboard score of 2.10 There is a part-2 notebook where deep learning has been attempted on kaggle GPU. Model architectures in that notebook consistently gave better leaderboard scores, such as, 2.05. However, overfitting dangers need to be checked through best practices. Part2: https://www.kaggle.com/code/sandeepmajumdar/abalone-age-prediction-deep-learning-3 (should be visible after I make the notebook public on kaggle)
9/5/2022
704 views
Tags:
#python
#machine-learning
#regression
#beginner