Qudsia Ahmad

Innovative Artificial Intelligence Engineer possessing strong analytical skills and detailed knowledge of machine learning evaluation metrics and best practices, producing models (e.g. "ClaimSights") with over 77% prediction accuracy. Offering over a year of hands-on experience in data analytics with expertise in predictive analysis, data mining, computational statistics and full-stack programming inclusive of dashboard. Logical and detailed professional with exceptional Python coding, in frameworks such as scikit-learn, pandas, numpy, matplotlib, Flask and functional scripting in JavaScript, using d3. Collaborative team player possessing first-rate problem-solving and collaboration capabilities.

ClaimSights (https://github.com/Yasmin-9/ClaimSights) : Machine Learning foresights into Health Insurance Claims, ClaimSights, a feature-rich full-stack application involving, data preprocessing analysis, ETL (postgres MySQL), back-end development (Flask, engine connection, scripting), front-end (JavaScript, HTML) and machine modeling with optimization techniques (SMOTE), engineered to predict a user's insurance claim status based on input fields like sex, age, and diagnosis code. The submitted values are processed through Flask's request method, initiating a query within the random forest machine model. This model accesses data stored in a MySQL database, retrieves the prediction with a 78% accuracy, and sends it back to the application, ultimately delivering the outcome to the user.

Provincial Crime Data during COVID-19 Dashboard (https://github.com/qudsia99/covid19-dashboard):  This full-stack project analyzes crime, employment, and income data of Canadians from 2018-2021. It explores the impact of employment changes and income on crime fluctuations, involving data cleaning, ETL, and backend development (Flask, locally built API). The frontend (HTML, Js, Chart.js, D3) visualizes the findings. Focusing on statistical analysis of coronary heart disease using methods like hypothesis testing, chi-tests, and p-values on data from the 'Framingham Heart Study.' Leveraging numpy, pandas, scipy.stats, and matplotlib, the study examines factors like cholesterol, smoking, BMI, and age, revealing a positive correlation between cholesterol levels and CHD development. The analysis includes assessing p-values and hypothesis testing to verify findings.