Mushroom Classification using Machine Learning

This project classifies mushrooms as edible (e) or poisonous (p) using supervised machine learning algorithms.

Problem Overview

Mushroom poisoning is caused by eating poisonous mushrooms mistaken for edible ones. This system aims to help identify mushrooms accurately to prevent such incidents.

Algorithms Used

Three machine learning algorithms were implemented and compared:

• Logistic Regression
• Gaussian Naïve Bayes
• Decision Tree Classifier

Tools and Libraries

• Python
• Jupyter Notebook
• Pandas
• NumPy
• Scikit-learn
• Matplotlib
• Seaborn

Dataset

• Source: UCI Machine Learning Repository (Mushroom Dataset)
• Preprocessing: handled null values, encoded categorical data, scaled features, and split data 80/20 for training/testing.

Performance Metrics

• Confusion Matrix
• Accuracy
• Precision
• Recall
• F1 Score
• AUC-ROC

Results Summary

Algorithm	Accuracy	Precision	Recall	F1-Score
Logistic Regression	93%	94%	90%	92%
Gaussian Naïve Bayes	~85%	86%	100%	92%
Decision Tree	~88%	89%	89%	89%

Best Model: Logistic Regression

Requirements

pandas
numpy
scikit-learn
matplotlib
seaborn
jupyter