Amna Faisal

Passionate Computer Science student specializing in NLP. Eager to apply academic knowledge and practical skills to innovative research projects in the field. Seeking a Ph.D. position in NLP research where I get opportunities to research and implement cutting-edge solutions, contribute to advancing the field, and make meaningful contributions. Sophisticated Machine Learning Engineer with background in independent research using intuitive, web-based architecture. Skilled in [Skill] with documented history of discovering methods to intelligently use data to enhance user experience. Effectively researches techniques for novel approaches to problems, develops prototypes to assess viability of approach and deploys application into production yielding insights to expand customer-consciousness.

SentimentScope: A Near Real-Time Dashboard For Public Emotion Monitoring Using Social Media Platforms

Topic of research for Masters which focuses on leveraging cutting-edge techniques in sentiment analysis, topic modeling, and text summarization to analyze social media data, to predict the onset of public riots or gatherings that pose a threat to national harmony. Through rigorous exploration and experimentation, I aim to contribute to proactive measures for maintaining social stability.

Taylor's School of Computer Science Student Management Dashboard

Single-handedly developed Django-based Student Management Portal for my university's Computer Science Postgraduate School. The dashboard helps manage, track,  and update the progress of new applicants, MS and PhD students, and Aluminis.

Election Prediction Using Social Media Sentiment Analysis and Bayesian Optimization

Single-handedly developed a Django-based application which displays up-to-date predictions for Pakistan's Elections. The model was based on Twitter sentiment analysis, past elections result, and provincial polling results. Leveraging data from these sources  and optimizing results with Bayesian optimization, it yielded constituency-level predicted election winners daily. The model achieved a 7/9 seat accuracy in Pakistan's September 2022 bi-elections.

Analysis of Reforestation Potential in Pakistan Using Machine Learning and Satellite Imagery

Final Year Research Project, whereby different machine learning models were implemented on satellite data of Pakistan, to find out areas that have potential to sustain forests. 

Football Player Detection

Single handedly, developed a football player detection and tracking program for a client using YOLOv3 and darknet. YOLO was used to especially allow the program to track players on low quality videos such as those obtained from street or school football games.

Garbage Classification using Machine Learning

 A dataset of labeled garbage images was obtained from Kaggle. It consisted of six categories of waste: cardboard, glass, metal, paper, plastic, trash. CNN and KNN algorithms were trained to classify these images to their respective categories, and their accuracies were compared. 

• 2020 - Remote Certification of Introduction and Programming with IoT Boards, with Pohang University of Science and Technology, South Korea.
• 2016 - Community Service at SOS Children's Village