cv

Visiting Student Researcher

Earlham Institute, Norwich Research Park

• Conducting MSc dissertation research on in silico prediction of pathogenic cell types and pathways in Crohn's Disease under supervision of Prof Irene Papatheodorou and Dr Gregory Wickham
• Integrating multi-omics data including bulk RNA-seq and single-cell RNA-seq from Crohn's Disease patient cohorts with clinical metadata to identify disease-progression trajectories
• Performing cell type deconvolution on bulk RNA-seq data using state-of-the-art algorithms including CIBERSORTx, MuSiC, and Bisque wrapped in the CATD pipeline
• Analysing single-cell RNA-seq data using the EISCA pipeline for quality control, normalization, clustering, and cell type annotation to identify pathogenic cell subtypes
• Developing machine learning models to integrate deconvolved cell type proportions, cell-type-specific gene expression signatures, and clinical parameters to predict disease biomarkers and progression trajectories
• Working with large-scale GEO datasets including GSE57945 (359 paediatric samples), GSE192786 (40 Crohn's biopsies), and GSE93624 (235 paediatric samples)

Honorary Research Assistant

Nixon Research Group, University of Liverpool

• Developed and evaluated TimeGPT for Influenza surveillance, achieving RMSE of 873.27 for 1-week forecast, outperforming traditional models (NumPy, Matplotlib, Pandas, Scikit-learn, and TimeGPT)
• Led seasonal pattern analysis for SFTS in South Korea using Complex EMD, harmonic regression (R²=0.848), and bicoherence analysis to identify multi-scale disease transmission cycles (Python, R, and Epidemiological Modelling)

Project Engineer

Asia Petroleum Limited

• Developed a hybrid ESN-LSTM model achieving R-squared of 0.975 and DA of 94.12% on long-short term forecasts of the KSE-100 Index (published in ICRAI 2024) (Python, NumPy, and Scikit-learn)
• Implemented real-time data visualization dashboards improving project delivery timelines by 20% (Power BI)

Data Science Intern

Additech Sim, University of Stuttgart

• Developed a Reinforcement Learning model which optimised pressure of shaft-hub connection 98% faster than the existing Q-learning approach (Python, Stablebaselines-3, OpenAI Gymnasium, NumPy, Pandas, and Scikit-learn)
• Applied and evaluated deep reinforcement learning using Python, StableBaselines-3, and OpenAI Gymnasium

Research Intern

PRL, National University of Sciences and Technology

• Led a team of 6 undergraduate students to design and develop a working prototype of power theft detection system (published in ICET 2023) (Python with Raspberry Pi, ThingSpeak, NumPy, and Pandas)
• Programmed Raspberry Pi for real-time transmission of voltage, current, and power factor readings to the ThingSpeak server

Masters

University of East Anglia

Data Science for Biology

• Data Science & Bioinformatics
• Statistics for Biologists
• Data Mining

Bachelors

National University of Sciences and Technology

Electrical Engineering

• Machine Learning (A)
• Digital Image Processing (A)
• AI & Decision Support Systems (B+)
• Numerical Methods (B+)
• Calculus & Analytical Geometry (A)
• Digital Signal Processing (B+)

Army Public School and Colleges System

GCE Ordinary and Advanced Level

Highest Achiever in Class of 2019

Army Public School and Colleges System

Graduated with school-level distinction

Runner-up in Hi-Robotec Robofiesta 2.0

NUST Robotics and Artificial Intelligence

Co-founder of NUST Robotics and Artificial Intelligence

Winner of Inter-College Cycling Championship

Army Public School and Colleges System

Served as the College Head Boy with active participation in sports

Attention-Based Deep Convolutional Neural Network for Classification of Generalized and Focal Epileptic Seizures

Epilepsy & Behavior

Taimur Shahzad Gill, Muhammad Ayaz Shirazi, Syed Sajjad Haider Zaidi. Epilepsy & Behavior 155.109732 (Feb. 2024)

Time Series Forecasting of KSE-100 Index Using a Hybrid ESN-LSTM Model

2024 International Conference on Robotics and Automation in Industry (ICRAI)

Taimur Shahzad Gill, Syed Ibrahim Zahid

Early Detection of Mesothelioma Using Machine Learning Algorithms

The 7th International Electrical Engineering Conference

Taimur Shahzad Gill, Muhammad Ayaz Shirazi, Syed Sajjad Haider Zaidi

IoT Based Smart Power Quality Monitoring and Electricity Theft Detection System

2021 16th International Conference on Emerging Technologies (ICET)

Taimur Shahzad Gill, Durr E Shehwar, Hira Memon, Sobia Khanam, Ali Ahmed, Urooj Shaukat, Abdul Mateen, Syed Sajjad Haider Zaidi

MHA-CNN for Classification of Generalised Epileptic Seizures

Proposed a novel CNN architecture with a multi-headed attention mechanism to classify the 5 types of generalised epileptic seizures.

• Achieved an average training accuracy of 99.1% and testing accuracy of 98.4% for multiclass seizure classification
• Technologies: Python, Tensorflow, Keras, MNE, PyEDF, NumPy, Pandas, and Matplotlib

KneeViT: Hybrid Architecture for Knee MRI Classification

Developed novel hybrid architecture combining VGG Transformer and OverLoCK ConvNet for knee MRI classification.

• Achieved AUC scores of 0.919 (abnormal), 0.809 (ACL tear), and 0.760 (meniscus Tear) with average of 0.845 after 50 epochs
• Technologies: Python, TensorFlow, PyTorch, NumPy, and Scikit-learn

Indigenous Development of a Low-Cost EEG Acquisition System

Developed a functional prototype of an EEG acquisition system to amplify the EEG signal 150 times.

• Carried out live testing of the prototype using Ambu EEG electrodes and visualised the signal using 12-bit ADC
• Technologies: C with Arduino IDE