A collection of data analysis and machine learning projects.
EDA of the highest-rated video games on Metacritic — all-time top 100 and the top 1,000 games from 2013–2022. Covers score distributions, critic vs user ratings, and platform trends over time. Data scraped with rvest.
R tidyverse rvest gtsummary
Classification model predicting heart attack occurrence based on clinical measurements (age, heart rate, blood pressure, CK-MB, troponin) collected during ambulance transport. Includes EDA and comparison of model features against clinical reference values.
R tidyverse plotly corrplot
Logistic regression model predicting passenger survival on the Titanic. Explores survival rates by class, sex and age, including a t-test confirming statistically significant age difference between survivors and non-survivors.
R tidyverse caret
Exploratory analysis of ~28,000 products from BigBasket, an Indian online grocery platform. Covers category share, rating distribution, and price/discount patterns across brands and categories.
R tidyverse
Fine-tuned YOLOv8 to detect 20 types of aircraft in satellite imagery. Trained and compared 4 model variants (YOLOv8n/m × frozen/unfrozen layers) on 3,821 images over 100 epochs using a Kaggle P100 GPU.
Python ultralytics YOLOv8
Course project covering time series modelling and forecasting methods in R.
R
Real-time vehicle detection and counting from an RTSP camera stream using object detection models (YOLOv3 via ImageAI, YOLOv8 via ultralytics).
Python ultralytics OpenCV