This repository, Sequential Machine Learning, is a personal project dedicated to my exploration and experiments with sequential data processing using advanced Machine Learning techniques such as Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM) networks, Gated Recurrent Units (GRUs), and Transformer models.
The main objective here is to build proficiency in handling sequential data, which is integral to many application areas, from Natural Language Processing (NLP) to stock forecasting. Over time, my aspiration is to apply these techniques to more complex robotics problems in the field of prediction and decision making.
Sequential data refers to the type of data where the order matters for interpretation. Examples of such data include text, speech, video, and time-series data such as stock prices. This project is my journey into the different techniques used in handling such data.
The technologies being used in this repository include:
- Python as the primary programming language.
- TensorFlow and PyTorch for building and training ML models.
- NumPy and Pandas for data manipulation.
- Matplotlib and Seaborn for data visualization.
- Bidirectional RNN for text classification on the AG_News dataset.
- Simple RNN from scratch used to predict IBM stock data.
In the near future, I plan to work on:
- Exploring different variants of RNNs including LSTMs and GRUs.
- Delving into attention mechanisms and Transformer models.
- Applying these techniques to more complex problems like machine translation and text summarization.
- Custom projects related to prediction and decision making in the context of robotics