Data Science · Operations Research · Mobility
I turn operational data into decisions.
I analyze operational data, build predictive models, and connect their outputs to optimization systems for routing, capacity, location, and scheduling.
DECISION GRAPH / 001 FEASIBLE
OBJECTIVE391expected cost
CAPACITY03slots / cycle
CONFIDENCE92%bootstrap wins
01 ROUTING×02 LOCATION×03 ASSIGNMENT×04 SIMULATION×05 ML → OPT
YS / ISTTHE PERSON BEHIND THE MODELS
Data Science,
with a decision bias.
I study Industrial Engineering at MEF University alongside a Data Science and AI minor. An Erasmus exchange at TU/e deepened my work in transport and mobility; today, I apply the same systems thinking to operational data, predictive models, and optimization.
Read the full profileBASEIstanbul
FOCUSDATA → DECISION
PROJECTS09 public
SELECTED WORK
Built, measured, questioned.
CASE / 01ML + OPTIMIZATION
MLOMAINTENANCE / MODEL
C1C2C3C4C5
RUL → SCHEDULE
YUSUF SAMI / OR LABFEASIBLE
Predictive maintenance that ends in a decision
Remaining Useful Life predictions connected to a maintenance scheduling model—not left as an isolated accuracy score.
- Python
- PyTorch
- XGBoost
- OR-Tools
TEST RMSE 17.21OPEN ↗
CASE / 02ROUTING
ALNSPDPTW / MODEL
D
P₁
D₁
P₂
D₂
YUSUF SAMI / OR LABFEASIBLE
Pickup and delivery with time windows
An Adaptive Large Neighborhood Search implementation for constrained, real-world routing decisions.
- Python
- ALNS
- Combinatorial Optimization
ALNSOPEN ↗