top of page

Spam Detection System using Recurrent Neural Networks

This project involved developing and training a character-level recurrent neural network (RNN) model using PyTorch to classify SMS text messages as spam or non-spam, achieving a validation accuracy of 96.5%.


I implemented data preprocessing and balancing techniques to handle class imbalances, significantly improving the model's performance in detecting minority class instances. Additionally, I optimized the model's hyperparameters and evaluated performance metrics, including false positive and false negative rates, to ensure robust spam detection capabilities.


Through this project, I demonstrated the power of RNNs in text classification tasks and highlighted the critical importance of thorough data handling and model evaluation for achieving high-performance machine learning solutions.


To learn more about this project, please refer to the file below.




bottom of page