<Hello World />
Albert Jojo
Crafting intelligent solutions through data-driven insights and machine learning.
Get to know me better
Pick a question — the answers come from me, not a model.
Skills
Tools and technologies I work with.
Languages
AI / ML
Cloud
DevOps
Database
Experience
Research and work.
Undergrad Student Researcher
- Co-authored a paper predicting Li-ion battery discharge capacity with ML algorithms
- Improved model performance by 40–50% through Optuna hyperparameter tuning with stratified k-fold cross-validation
- Generated foundational datasets by extracting critical performance metrics from literature using ImageJ
Undergrad Student Researcher
- Supported a published research paper by inspecting and labeling hundreds of solar panel defect images
- Built structured databases with Python and SQL pipelines to enhance data accessibility
- Spearheaded database–cloud inter-connectivity, reducing dataset preparation turnaround time
Projects
Things I've built and explored.
Real-Life Pokedex
Image classification app that identifies Pokemon by region using EfficientNet-B0 trained on a custom-scraped dataset. Includes a web frontend and Docker containerization.
CSIRO-Biomass
Ensemble model for biomass prediction using self-supervised ResNet-50 image embeddings and XGBoost regression, optimized for small labeled datasets.
S&P 500 Trading Strategy
Tactical trading strategy using time-series feature engineering and XGBoost with time-aware validation for risk-aware portfolio allocations.
Baseball Analytics
Analyzed defensive alignment patterns at Black & Gold Analytics. Predicted optimal field setups using XGBoost and Random Forest.
Let's Connect
Open to research collaborations, internships, and conversations about data science and ML.