Muhammad Rizwan
Associate Lecturer
Ext:
Muhammad Rizwan holds an MS in Data Science from NUCES and a BS in Computer Science from Superior University, Lahore. He is currently an Associate Lecturer at the University of Central Punjab, bringing his expertise in AI, machine learning, and data analytics to the classroom. Previously, he worked as an Associate Data Scientist at Reallytics.ai, where he developed innovative solutions using tools like ChatGPT and DeepFace. His projects, such as REALTALK and ATTENDANCE SYSTEM, highlight his technical skills. An active researcher with publications under review, Muhammad has presented at international conferences like icSoftComp2023. Proficient in Python, SQL, and data analysis tools, he is certified in machine learning and data science, making him a valuable contributor to academia and industry.
| MS in Data Science | Machine Unlearning | FAST, National University of Computer & Emerging Sciences | 2024 |
| BS in Computer Science | Mobile Application | The Superior University | 2022 |
| Associate Data Scientist | Reallytics.ai | 2023-2024 |
| Machine Learning Instructor | PNY Trainings | 2024 - 2024 |
| Teacher | Al-Noor Science Academy | 2014-2022 |
Click to View
Machine Learning, Deep Learning, Computer Vision, NLP, Data Science, Big Data Analytics
Federated Learning, Machine Unlearning and Person Re-Identification
Click to View
REALTALK
- Analyze call transcripts to identify the top-performing agent monthly using performance metrics
- Implement Chat-GPT prompt engineering for agent identification, diarization, call summary extraction, and extraction of top 5 topics with guidelines. Agent emotions contribute to monthly performance analysis
DEMAND FORECASTING H&M
- The garment store aims to accurately predict the monthly demand for its products to optimize inventory management
- Ten distinct models were trained using Darts for a univariate problem, and performance evaluation relied on RMSE metrics to assess predictive accuracy
SILVER TREE BRANDS
- Use Google Big Query to fetch the data, do some transformation by taking guidance from KNIME Legacy workflows
- Make the connection between Dataform and Write the SQLX scripts for performing the transformation on the data, and save the results to the BQ datasets
ATTENDANCE SYSTEM
- Develop an AI-based attendance system for efficient and automated tracking of personnel attendance
- The DeepFace library facilitates face recognition, with recognized faces triggering the upload of payloads to an S3 bucket. Subsequently, a Lambda function is triggered to store the data in DynamoDB
Click to View
- [2024] Muhammad Rizwan, Aamir Wali, Muhammad Saif-ul-Islam, Sana Fatima, 2024, Unlearn at Once, Machine Learning, [Under review]
- [2024] Muhammad Rizwan, Hamza Naveed, Rana Muhammad Zulqarnain, Comparative Analysis of Neural Network Architectures for Image Classification,2024,[Under Review]
- [2024] Sumaira Rafique, Muhammad Rizwan, Towards Reliable Federated Conditional GANs Mitigating Label Flipping Attacks Through Loss-Based Detection, 2024,
- [In Progress] [2024] Muhammad Rizwan, Hamza Naveed, Unraveling the Challenges of Urdu Script Recognition: A Deep Learning Approach Using the PSL Dataset,2024, [In Progress]