Support-Vector-Machines-for-Spam-Detection

This repository contains a Jupyter Notebook that demonstrates the implementation of Support Vector Machines (SVM) for spam detection. The notebook includes data preprocessing, feature extraction, training the SVM model, and evaluating its performance.

Files

• Support Vector Machines.ipynb: The Jupyter Notebook containing the implementation of SVM for spam detection.

Requirements

• Python 3.x
• Jupyter Notebook
• NumPy
• Matplotlib
• SciPy
• scikit-learn
• NLTK
• stemming

Usage

1. Clone the repository:

   git clone https://github.com/cizodevahm/Support-Vector-Machines-for-Spam-Detection
2. Navigate to the repository directory:

   cd Support-Vector-Machines-for-Spam-Detection
3. Open the Jupyter Notebook:

   jupyter notebook Support Vector Machines.ipynb

Implementation Details

• Data Preprocessing: The notebook includes functions to preprocess emails by converting them to lowercase, removing HTML tags, replacing numbers, URLs, email addresses, and dollar signs with placeholders.
• Feature Extraction: The notebook extracts features from emails by tokenizing, stemming, and mapping words to a vocabulary list.
• Training the SVM Model: The SVM model is trained using a linear kernel on a dataset of spam and non-spam emails.
• Evaluating Performance: The notebook evaluates the model's performance by calculating training and test set accuracy and identifying the top predictors for spam.

License

This project is licensed under the MIT License.