Price Predicting Transformer
- Category: Machine Learning
- Tools & Frameworks: NumPy, PyTorch
- Project Duration: Feb-March, 2024
Details
- Developed a Transformer-based model to predict stock prices by capturing intricate patterns and long-range dependencies in financial time series data.
- Designed and implemented a Transformer model with multiple attention heads and feedforward layers, enabling the capture of complex dependencies in data.
- Trained the model on a substantial dataset with over 21 million sequences with varying sequence lengths.
- Gained expertise in Transformer architecture and self-attention mechanisms, enhancing the ability to model complex sequential data.