🔥 Sequence-to-Sequence Event Prediction

🚀 An advanced deep learning-based sequence-to-sequence (Seq2Seq) model for event prediction with cross-validation, hyperparameter tuning, and robust evaluation.

🔗 GitHub Repository: Sequence-to-Sequence Event Prediction

---

📌 Project Overview

This repository implements a sequence-to-sequence (Seq2Seq) model using LSTM networks to predict event sequences based on historical data. It supports:

• K-Fold Cross-Validation for model robustness.
• Hyperparameter tuning to find the best configuration.
• Post-training analysis to select the best model.
• Retraining & Evaluation on unseen test data.

---

🏗 Project Structure

📂 Sequence-to-Sequence-Event-Prediction
│── main.py              # Runs the full training pipeline
│── train.py             # Training & validation functions
│── post_analysis.py     # Selects the best model after training
│── retrain.py           # Retrains the best model & evaluates on test set
│── model_definition.py  # LSTM-based Encoder-Decoder model
│── data_loader.py       # Data preprocessing & k-fold dataset preparation
│── metrics.py           # Custom loss & evaluation metrics
│── utils.py             # Utility functions (logging, config handling, plotting)
│── config.yaml          # Configuration settings for training jobs
│── Events.csv           # Dataset file (input)
│── README.md            # Project Documentation

---

⚙️ Installation & Setup

1️⃣ Clone the Repository

git clone https://github.com/ShalinVachheta017/sequence-2-sequence-Event-Prediction.git
cd sequence-2-sequence-Event-Prediction

2️⃣ Install Dependencies

pip install -r requirements.txt

3️⃣ Ensure Correct Folder Structure

Ensure that Events.csv (dataset file) is present in the root directory.

---

🚀 How to Train the Model

The training pipeline performs:

• Data Loading & Preprocessing
• K-Fold Cross-Validation
• Hyperparameter Optimization
• Performance Metrics Computation

Run the Training Script

python main.py

This script:

1. Loads data from Events.csv
2. Trains the model with multiple hyperparameters (specified in config.yaml).
3. Saves training logs and model outputs in ./logs and ./outputs directories.

---

📊 Post-Training Analysis

To analyze the trained models and select the best-performing one:

python post_analysis.py

This script selects the best model based on validation coverage and accuracy while avoiding overfitting.

---

🔄 Retraining and Final Evaluation

Once the best model is selected, it can be retrained multiple times and evaluated on test data.

python retrain.py

This script:

1. Loads the best model configuration.
2. Retrains it multiple times for robustness.
3. Evaluates on the test set.
4. Saves test performance results in final_test_results.csv.

---

🛠 Customization

To modify the training configurations (batch size, learning rate, epochs, etc.), edit config.yaml:

general:
  batch_size: 8
  epochs_list: [40, 80, 100, 120]
  learning_rates: [0.001]
  hidden_sizes: [64, 128, 256]
  num_layers_options: [1, 2, 3]
  bidirectional_options: [true, false]

---

📈 Performance Metrics

The model is evaluated using:

• Cross-Entropy Loss for training and validation.
• Accuracy & Coverage for post-training and evaluation.

The results are stored in:

• all_jobs_results.csv (training phase)
• final_test_results.csv (final evaluation)

---

🤖 Model Architecture

The model is a Bidirectional LSTM-based Encoder-Decoder:

Encoder (LSTM) → Decoder (LSTM) → Fully Connected Layer

The decoder uses teacher forcing with a ratio of 0.3 during training.

---

🏆 Results & Visualization

• Training loss curves, validation loss curves, and accuracy are saved in ./outputs/

---

📬 Contributing

Feel free to fork, open issues, or submit pull requests! 😊

---

📝 License

This project is open-source under the MIT License.