Classification with Playground Series - S4E8
Kaggle Playground August 2024 - Binary Prediction of Poisonous Mushrooms
Goal: The goal of this competition is to predict whether a mushroom is edible or poisonous based on its physical characteristics.
Evaluation: Submissions are evaluated using the Matthews correlation coefficient (MCC).
Learnings:
- Define unconventional metrics
- Handling noises in data
- Consolidate ML workflow : train-test split > feature engineering + preprocessing > train + fit models > evaluate > fine-tune models
- Deep Learning
Table of Contents
1. Data Preparation and Cleaning
2. Exploratory Data Analysis (EDA)
3. Feature Engineering (Frequency Encoder, Label Encoder)
4. Machine Learning (XGBoost, LGBM, CatBoost)
5. Deep Learning
6. Ensemble
7. Evaluate
8. Model Evaluation
9. Predict & Submit
Achievement: My MCC score is 0.98
* Check out more details in my Medium blog.
