Share Profile
MUHAMMAD MUNAWWAR ANWAR
Aspiration Statement
I intend to pursue postgraduate studies in Data Science, with a focus on data-driven decision-making. I am inclined towards data analytics and research as a career option. My other passions include travelling and gardening.
Core Skills
- Python
- LaTex
- SQL (Structured Query Language)
- C++
- C#
- JavaScript
Academic Awards / Achievements
- Dean's List, Fall 2020
- High Achievement Academic Scholarship Fall 2020
Experience
Leadership / Meta-curricular
- Co-lead, Database Module — Code Play 4.0 Team Member, Mind Maze Module — OMICRON V
Internship / Volunteer Work
- Computer Science Research Assistant - Habib University
- IT Intern — Ismail Industries Limited (July 2021— August 2021)
- C# Intern — Logistic Edens (July 2021)
Final Year Project
Project Title
Compression based Perceiver Architecture
Description
The FYP is a research project based on Google DeeplVind's Perceiver Architecture, which is a step toward Artificial General Intelligence. The Perceiver Architecture is a general architecture that can work for multiple modalities such as audio, video, images. This project aims to reduce the memory and computational power bottleneck when training the Perceiver Architecture. The hypothesis that we are testing is if the learned representation is separable enough, the Perceiver should provide accuracy similar to the raw data. Autoencoders and contrastive learning techniques are used to learn representations of the data. The initial results from the prototype are encouraging, however, the model needs to be trained on a larger dataset.