Forest Cover Type Classification Model

.This project predicts the forest cover type based on cartographic and environmental features

:It applies Machine Learning models such as

Random Forest (baseline)-

XGBoost (with hyperparameter tuning)-

-----------------------------------------------

Project Workflow:

Data Exploration & Cleaning – inspected data types, missing values, and separated continuous vs one-hot features-

Feature Preparation – encoded categorical variables (Wilderness Area, Soil Type)-

Model Training – applied Random Forest and XGBoost-

Evaluation – compared Precision, Recall, F1-Score, Accuracy, and Cross-validation result-

-----------------------------------------------

Results:

- Random Forest → Accuracy: ~0.954 | Strong baseline with stable performance & clear feature importance.

- XGBoost (tuned) → Accuracy: ~0.973 | Outperformed Random Forest with higher F1, Precision, and Recall.

XGBoost was the final choice, delivering the most accurate predictions.

-----------------------------------------------

Business Impact:

With highly accurate cover type predictions, XGBoost reduces classification errors in ecological and forestry applications. This allows decision-makers to rely less on manual surveys and make faster, data-driven environmental assessments.

-----------------------------------------------

Tech Stack:

Python | Pandas, NumPy | Matplotlib, Seaborn | scikit-learn, XGBoost