• 🔭 Current Focus: Building knowledge in Data Science, ML & AI

• 🌟 Background: 7+ years in Game Engine/Middleware development. Mainly focusing on mobile platforms, user-facing UI and API development, data-heavy internal tool development.

• 🌱 Interests:

  • Working on user-centric, data-driven applications
  • Applying DS & ML to solve real-world problems

	Python
		Traditional ML (Scikit-learn, XGBoost, LGBM etc.)
		Other Python

	iOS / Obj-C
	C# / .NET / Mono
	C++
	Unity3D

	React / Node.js
	Linux / Docker / etc.
	VCS / Git / Mercurial / Perforce

AgeGenderClassifier

Building a MobileNet age and gender classification model using the UTKFace dataset. This model achieves comparable performance to significantly more complex models like SENet, ResNet, and VGG. The project includes:

• Extensive manual data analysis
• Tuning using Weights & Biases
• Augmentation-based oversampling and synthetic sample generation

HomeCreditDefaultRisk

Developing LGBM and XGBoost models for predicting credit default risk. Key features:

• Extensive feature engineering using Feature Tools/DFS
• Hypothetical risk model and ROI analysis

WashingtonPostPoliceShootings

• Jupyter Notebook/Presentation with exploratory data analysis on Police Shootings Database 2015-2023, Washington Post

FootballBettingModel

• Building LGBM, XGBoost, and other models for predicting football match outcomes
• Includes extensive EDA and feature engineering (individual player analysis, rolling performance metrics, etc.)