SPAM-CLASSIFIER

Data Science: Email Classifier

This project involves building an email classifier using the Multinomial Naive Bayes algorithm to classify emails as spam or non-spam (ham). It fetches the Spambase dataset from the UCI Machine Learning Repository, splits it into training and testing sets, trains a Multinomial Naive Bayes classifier, and evaluates its performance using various metrics.

Project Overview

The email classifier project includes the following steps:

1. Data Fetching: Fetching the Spambase dataset from the UCI Machine Learning Repository using the fetch_ucirepo function.

2. Data Preprocessing: Splitting the dataset into features (X) and target labels (y), and further dividing them into training and testing sets using the train_test_split function from sklearn.model_selection.

3. Model Training: Training a Multinomial Naive Bayes classifier using the MultinomialNB class from sklearn.naive_bayes.

4. Model Evaluation: Evaluating the performance of the classifier using accuracy scores, confusion matrix, and ROC curve.

Setup

To set up the project environment, ensure you have the necessary dependencies installed. You can install them using the following command:

pip install numpy pandas scikit-learn matplotlib ucimlrepo



## Setup

To execute the email classifier code, run the provided Python script (email.py) in your preferred Python environment. Ensure that you have the required datasets accessible.


## Output

Upon execution, the script will output the following:

Train accuracy: The accuracy of the classifier on the training set.
Test accuracy: The accuracy of the classifier on the testing set.
Confusion matrix: A matrix showing the true positives, true negatives, false positives, and false negatives.
ROC curve: A graphical representation of the Receiver Operating Characteristic (ROC) curve.
## License

This project is licensed under the MIT License. See the LICENSE file for details.