SADIA TABASSUM, PhD

Machine Learning & Data Science Specialist

Find Out More

About Me

  • ➤ I earned my PhD in Computer Science from the University of Birmingham. I specialise in the intersection of Machine Learning, Data Science, and Software Engineering to develop innovative solutions for real-world challenges. I have 5+ years of practical experience using Python, Machine Learning, and Data Science tools and libraries.
  • ➤ Skills:
    • Programming languages: Python, Java, Matlab, SQL, Postgresql
    • Data Analysis: Pandas, Numpy, Scipy
    • Data Visualisation: Plotly, Seaborn, Matplotlib
    • Machine Learning: Supervised Learning, Classification, Regression, NLP, Ensemble Methods, Optimisation
    • Tools & Libraries: Scikit-Learn, SPSS, PyTorch, Flask, AWS, Excel
    • Other: Probability, Statistics
  • ➤ Training:



Go to Projects

Recent Projects

Selected projects in data science, machine learning and NLP

Stack Overflow Developer Survey Data Analysis

Data Science | Data Analysis

View code on Github | Read Blog

I performed a thorough data analysis on the Stack Overflow Annual Developer Survey Data to uncover insights about data scientists. I addressed specific research questions through my analysis and developed a machine learning model to predict the salaries of data scientists.

Static Badge Static Badge Static Badge

Disaster Response Pipeline

NLP | Text Classification | ETL Pipeline

View code on Github

I have created an ETL pipeline, ML pipeline and Web application based on Flask to categorise real-time messages during disaster events. I have applied NLP techniques to process the text data. The dataset is provided by Figure Eight in collaboration with Udacity.

Static Badge Static Badge Static Badge Static Badge

Just-In-Time Software Defect Prediction

Machine Learning

View code on Github | Read Paper

In this project, I have developed three different novel online Machine Learning approaches for real-time software defect prediction. These models demonstrated enhanced predictive performance, with G-Mean improvement reaching up to 48.16% during concept drift periods.

Static Badge Static Badge

Online Hyper-Parameter Tuning

Machine Learning

I have proposed and implemented a novel online hyper-tuning algorithm capable of tuning machine learning model hyper-parameters in real-time. This novel method can regularly identify optimal hyper-parameter combinations, minimizing declines in the machine learning model's predictive performance.

Static Badge Static Badge Static Badge Static Badge

View GitHub View LinkedIn Read My Blog