I will build a custom machine learning model

I build Machine Learning and Deep Learning solutions using Python turning your raw data into accurate, real-world predictions and intelligent systems.

My core stack: XGBoost, Random Forest, Scikit-learn for prediction models, and YOLOv8 with OpenCV for real-time computer vision and object detection tasks.

Recent work: AsteroidWatch, a real-time asteroid hazard prediction system built using 5 NASA APIs, trained on 38,573 records with XGBoost and SMOTE for imbalanced data prioritizing high recall to minimize missed hazards, a safety-first approach over just chasing accuracy numbers.

What I offer:

Data cleaning and feature engineering

Imbalanced dataset handling (SMOTE, class weighting)

ML model training and evaluation (XGBoost, Random Forest)

Deep Learning and object detection (YOLOv8, OpenCV)

Clear performance reports (accuracy, precision, recall, confusion matrix)

Optional deployment as a live Streamlit web app

Live demo: asteroidwatch-v2o.streamlit.app

Let's turn your data into predictions you can actually trust.