Evolving strategy candidates
I explored genetic programming with DEAP to generate trading formulas and evaluate candidates using NumPy/Pandas backtests. I later restored a baseline strategy and improved state carry in the backtest kernel.
LAB / INDEPENDENT / RESEARCH / 2026.02
Epic Invest — Genetic Strategy Research
PROJECT I explored genetic programming to generate crypto strategy formulas and assess candidates through cost-aware historical backtests.
SKILLS Python · DEAP · NumPy · Pandas · Backtesting
HIGHLIGHT Strategy evolution and cost-aware backtests; research results only.
OVERVIEW
About the project
Explores crypto strategies with genetic programming. Generates and evaluates candidate trading formulas. Runs cost-aware backtests for research.
DEVELOPMENT JOURNEY
How the work evolved
Research and evaluation
I keep research hypotheses, historical simulations and paper-mode results distinct from actual trading outcomes.