IMDb Sentiment Analysis

Overview

This project implements and compares classical Machine Learning and Deep Learning models for binary sentiment analysis on the IMDb Large Movie Review Dataset using Scikit-learn and PyTorch.

The project evaluates traditional machine learning algorithms alongside recurrent neural networks (RNNs) to analyze their performance on the same sentiment classification task. Performance is assessed using Precision, Recall, F1-score, and learning curves.

Implemented Models

Classical Machine Learning

• Naive Bayes
• Logistic Regression
• Random Forest

Supporting Components

• Text Preprocessing
• Feature Extraction
• Learning Curve Evaluation

Deep Learning

• Bidirectional RNN
• LSTM/GRU
• PyTorch implementation
• Model training and evaluation

Dataset

• IMDb Large Movie Review Dataset
• Binary sentiment classification (Positive / Negative)

Technologies

• Python
• Scikit-learn
• PyTorch
• NumPy
• Matplotlib

Project Structure

.
├── main.py
├── preprocessing.py
├── feature_extraction.py
├── naive_bayes.py
├── logistic_regression.py
├── imdb_rf_tuned.py
├── imdb_rnn.py
├── evaluation.py
├── learning_curves.py
├── create_loss_plot.py
├── learning_curves_rf.png
├── vocabulary.pkl
├── best_model.pt
├── requirements.txt
├── LICENSE
└── README.md

Installation

pip install -r requirements.txt

Run

Classical Machine Learning:

python main.py

Deep Learning:

python imdb_rnn.py

Performance Evaluation

The implemented models are evaluated using:

• Precision
• Recall
• F1-score
• Learning Curves

The objective is to compare the effectiveness of classical Machine Learning algorithms with Deep Learning approaches for sentiment analysis.