I built a cross-basin well-log facies classifier using Random Forest and Scikit-learn, trained on integrated SEG 2016 and FORCE 2020 datasets. With Leave-One-Well-Out cross-validation, it achieved an average F1-score above 0.85, and I added a Streamlit application for data preprocessing, training, prediction, evaluation, and visualization.
For another project, I developed CWSF, a Scikit-learn-compatible ensemble classifier that combines Random Forest, Logistic Regression, and K-Nearest Neighbors using confidence weights based on Shannon Entropy. Iām also pursuing a Diploma in Data Science and Programming at IIT Madras and applying machine learning to geology, geophysics, and Earth sciences.

