Dr. Fee Predictor

Developed an ML model to forecast fees of Pakistani doctors using data from Marham.pk, deployed on Vercel Cloud. Explored various algorithms, finding LightGBM most effective; optimized its hyperparameters for reliability. Use cases span healthcare, medical consultations, tourism and insurance fields

Heart Attack Risk Prediction

Developed a precise model to predict heart attack risk among individuals based on several factors. The risk was categorized as High, Medium, or Low. Used ensemble Machine learning and tuned the hyperparameters of different models and found Random Forest Classification to be the optimized one.

Urban Traffic Prediction Model

Developed a Classification ML model to predict the flow of traffic in Tier-I and Tier-II cities based on multiple factors. Clustered the peak hours and vehicle count. Found the XGBoost to be the best algorithm for Traffic prediction. Tuned the hyperparameters of XGB and applied voting classification

Material Hardness Prediction Model

I employed Hist GradientBoost Regressor, specifically trained on the metadata, to identify the hardness of material on Mohs scale. This identification is based on Euclidean distance. The model is generic and can be utilized anywhere

Food Trends Analytics(Zomato)

The analytics project delves into the trend analysis of Popular Indian food platform Zomato. I have analyzed the food trends descriptively, predictively and Prescriptively covering all aspects of people flavours.