🎬 ANNImdbClassifier

A neural network-based classifier for IMDB movie reviews sentiment analysis using Keras.

📝 Description

ANNImdbClassifier is a Python implementation of a sentiment analysis model that classifies IMDB movie reviews as either positive or negative. The project uses a simple Artificial Neural Network (ANN) built with Keras to achieve this classification task.

✨ Features

• Loads and preprocesses the IMDB dataset from Keras
• Transforms text reviews into numerical vectors using one-hot encoding
• Implements a sequential neural network with multiple dense layers
• Visualizes training and validation metrics (loss and accuracy)
• Provides a clean, object-oriented implementation

🔧 Prerequisites

• Python 3.x
• Keras
• TensorFlow (backend for Keras)
• NumPy
• Matplotlib

🚀 Setup Guide

1. Clone the repository:

   git clone https://github.com/corticalstack/ANNImdbClassifier.git
   cd ANNImdbClassifier

2. Install the required dependencies:

   pip install tensorflow keras numpy matplotlib

💻 Usage

Simply run the main script:

python main.py

This will:

1. Load the IMDB dataset
2. Preprocess the data
3. Build and train the neural network model
4. Display training and validation metrics graphs

🏗️ Architecture

The classifier is implemented as a single class ImdbClassifier with the following methods:

• __init__(): Initializes the classifier, loads and preprocesses data
• build_model(): Creates and trains the neural network model
• load_data(): Loads the IMDB dataset from Keras
• decode_review(): Converts encoded reviews back to human-readable text
• vectorize_sequences(): Transforms integer sequences into binary matrices

The neural network architecture consists of:

• Input layer: Dense layer with 16 neurons and ReLU activation
• Hidden layer: Dense layer with 16 neurons and ReLU activation
• Output layer: Single neuron with sigmoid activation for binary classification

📊 Model Performance

The model visualizes both training and validation metrics:

• Loss over epochs
• Accuracy over epochs

These visualizations help in understanding the model's learning process and identifying potential overfitting.

📄 License

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

🔗 Resources

• Keras Documentation
• IMDB Dataset
• Neural Networks for Sentiment Analysis