Projects

Projects @ Data Machine Elite

Duration: 8 Months (Nov 2024 - July 2024)
Code: [link]
Description: Welcome to the Noor_LLM Projects repository! This repository will host various projects and experiments related to Language Model (LLM) development and applications. Stay tuned for updates and new projects in the field of Natural Language Processing and Artificial Intelligence.

Duration: 6 Months (Jan 2024 - July 2024)
Code: [link]
Description:This project introduces a new supervised learning method, Multi-Scale Convolutional LSTM with Transfer Learning (TL), to detect unusual activity in cellular networks.
Outcome: The project develops a DL based framework to detect anomalies in cellular networks. By combining ConvLSTM for spatial-temporal pattern recognition and transfer learning for leveraging pre-trained knowledge, the model enhances detection accuracy and generalization. The result is a robust and scalable anomaly detection system that improves cellular network reliability and security.

Duration: 6 Months (Jan 2023 - July 2023)
Code: [link]
Description: This project, developed by Nooruddin Noonari, focuses on implementing a Diffusion Probabilistic Model (DDPM) with a custom beta schedule, allowing dynamic adjustment during the diffusion process. The model architecture is based on a U-Net structure, with ResNet and Attention mechanisms incorporated for enhanced learning. This project serves as an initial step towards improving and customizing Diffusion Models.
Outcome: The initial results of the model are, representing the generated images after training. The results indicate a good starting point, with potential for further improvements.

Duration: 2 Years (Jan 2019 - Dec 2021)
Code: [link]
Description: This project is titled Indoor Positioning System using GAN and was developed by Nooruddin Noonari as part of the Master's in Software Engineering program at Tianjin University, Tianjin, China.
Outcome: The Indoor Positioning System (IPS) leverages a Generative Adversarial Network (GAN) to improve the accuracy and robustness of location-based services within indoor environments. Traditional WiFi-based positioning systems often struggle with issues like signal fluctuation and the high cost of building extensive fingerprint databases. This project addresses these challenges by using GANs to generate synthetic WiFi fingerprints, which enhance the system’s ability to predict positions accurately even in challenging environments.

Duration: 1 Year (Jan 2014 - Dec 2015)
Code: [link]
Description: This project is titled Sign Language Interpreter for Mute People and was developed by Nooruddin Noonari as a Final Year project at Sukkur IBA University under the Bachelor of Science in Computer Science program.
Outcome: The Sign Language Interpreter is a system designed to bridge the communication gap between mute/deaf individuals and those who can speak and hear. The system takes input in the form of Pakistan Sign Language (PSL) gestures and converts them into both audible and readable text formats. This allows a mute/deaf person to communicate more effectively with others.

Duration: 6 Months (Jan 2013 - July 2014)
Code: [link]
Description: This collection is designed to provide a strong foundation in Python programming, covering everything from basic concepts to advanced applications.
Outcome: These exercises were developed during my Bachelor of Science in Computer Science at Sukkur IBA University, where I gained comprehensive knowledge in various programming languages, including Python, and worked on impactful projects like a Sign Language Interpreter for Mute People.