Welcome to my Kaggle Notebooks repository! This repository contains Jupyter notebooks from various Kaggle competitions, where I explore data, build models, and analyze results. Each competition folder also includes a markdown subfolder containing the code and output exported as markdown for easy viewing and sharing.
- Goal: Classify body-focused repetitive behaviors (BFRBs) vs. non-BFRB gestures using multimodal wrist sensor data.
- Code:
timeseries_classification_CMI - Approach: Time series modeling with the
sktimelibrary, ensemble of classifiers, handedness normalization, and sensor-specific preprocessing.
- Goal: Develop predictive models for post-hematopoietic cell transplantation (HCT) survival while addressing equity concerns.
- Notebook:
transplanting_data_into_insights.ipynb - Approach: Data preprocessing, exploratory data analysis (EDA), feature engineering, and machine learning models.
- Goal: Predict listening times of podcast episodes.
- Notebook:
eda-ensemble-pytorch-xgboost-help-wanted.ipynb - Approach: Data preprocessing, exploratory data analysis (EDA), feature engineering, and machine learning models (XGBoost & PyTorch).
- Goal: Predict how many calories where burned during a workout.
- Notebook:
burining-calories-a-model-comparison-ridge-ensem.ipynb - Approach: Data preprocessing, exploratory data analysis (EDA), feature engineering, and machine learning models (various base models and ridge ensemble).
- Add more notebooks from other Kaggle competitions.
- Improve documentation and model explanations.
- Experiment with different machine learning techniques.
This repository is open-source and available under the MIT License.
📩 Contributions & Feedback: Feel free to fork the repository, open issues, or suggest improvements!