# Fungal Habitat Prediction Using Taxonomic Classification and Observation Data

**Technologies:** Python, NumPy, scikit-learn, Random Forest, LightGBM  

## Objective
To predict fungal habitat types based on taxonomic classification and observation data, aiding in conservation efforts and understanding ecological patterns.

## Key Contributions
- Applied Random Forest and LightGBM algorithms to build predictive models for fungal habitat classification.
- Conducted data preprocessing and standardization to enhance model accuracy and interpretability.
- Analyzed model performance to identify strengths in predicting specific habitat types, focusing on metrics like precision, recall, and AUC scores.

## Outcome
Achieved moderate accuracy scores with **Random Forest** at 56.41% and **LightGBM** at 57.96%, demonstrating the models' utility in predicting habitat types such as 'coniferous woods' and 'other.' This project serves as a foundation for future studies in ecological data and conservation.

[GitHub Repository for Fungal Habitat Prediction Using Taxonomic Classification and Observation Data](https://github.com/mrw-soumik/Fungal-Habitat-Prediction-Using-Taxonomy-Data)