I study how adversarial dynamics and execution constraints shape behavior in decentralized markets.
My work focuses on machine learning, MEV, market microstructure, and simulation-based analysis.

SSRN Working Paper
Paper

A Bayesian HMM–LSTM framework for multi-horizon regime detection with full backtesting, Monte Carlo stress tests, and diagnostics.

Repo: https://github.com/lucaskemper/mev-portfolio-analysis

Quantifies how sandwich attacks and gas regimes distort execution costs in AMMs.
Empirical basis for my MSc thesis on MEV-aware reinforcement learning.

ML/AI: PyTorch, TensorFlow, scikit-learn, hmmlearn
Systems: Docker, async pipelines, WebSockets
Data: NumPy, pandas, statsmodels
Languages: Python, R, MATLAB, SQL

• MEV and adversarial blockchain dynamics
• Decentralized market microstructure
• Machine learning for execution and state detection
• Simulation environments for multi-agent systems
• Reinforcement learning in financial & blockchain settings

Email: lucas@singularitytrading.io LinkedIn: https://linkedin.com/in/lucaskemper