Task was to improve rock type prediction by using labeled core and well log data for one oil and gas field with multiple horizons. We tried using Random Forest, XGBoost, Logistic Regression and SVM. Of all algorithms, when tuned Random Forest gave us the best results.
kovelja009 / eestec_ml_hackathon Goto Github PK
View Code? Open in Web Editor NEWTask was to improve rock type prediction by using labeled core and well log data for one oil and gas field with multiple horizons. We tried using Random Forest, XGBoost, Logistic Regression and SVM. Of all algorithms, when tuned Random Forest gave us the best results.