Skip to content

Latest commit

 

History

History
39 lines (29 loc) · 2.73 KB

index.md

File metadata and controls

39 lines (29 loc) · 2.73 KB
layout
default

About Me

Hi there. I am a Machine Learning guy, finishing my PhD at Imperial College London under the supervision of Abbas Edalat. My current research topic can be very briefly summarized as "Hierarchical model-based deep reinforcement learning for trading".

Research Interests

My other research interests now are in the domain of Deep Reinforcement Learning. I am interested in simple methods for non-parametric model-based reinforcement learning, semantic memory building from episodic memory using statistical properties of the environment, arbitrary hierarchical structuring of environments for planning, auxiliary reward functions, etc.

You can check out my longer story in the About section.

Publications

  1. Wiering, M. A., Schutten, M., Millea, A., Meijster, A., & Schomaker, L. R. B. (2013). Deep support vector machines for regression problems. In Proceedings of the international workshop on advances in regularization, optimization, kernel methods, and support vector machines: theory and applications.
  2. Millea, A. (2014). Explorations in echo state networks. Department of Artificial Intelligence, Groningen University, MA, Netherlands. pdf cite
  3. Using Deep Reinforcement Learning with Hierarchical Risk Parity for Portfolio Optimization https://www.mdpi.com/2227-7072/11/1/10/pdf
  4. Deep reinforcement learning for trading—A critical survey https://www.mdpi.com/1361204

Unpublished work

  1. A.Millea, D.Podareanu, M.A.Wiering: Optimal action selection for the Foreign Exchange market using a Sarsa-Echo State Network pdf cite
  2. A.Millea, A.Edalat: Natural Boltzmann Machines [pdf](Natural Boltzmann Machines.pdf) cite
  3. A.Millea: Message passing algorithms for time series [pdf](Message passing algorithms for time series.pdf) cite
  4. A.Millea: Information Geometry for Machine Learning pdf cite
  5. A.Millea: A deep reinforcement learning short survey pdf cite
  6. A.Millea: Efficient long term prediction of pixels with the Modular Autoencoder for Atari games pdf cite
  7. A.Millea: Efficient hierarchical subgoal discovery for reinforcement learning [pdf](Efficient hierarchical subgoal discovery - draft.pdf) cite
  8. A.Millea: A cognitive architecture based on Deep Reinforcement Learning pdfcite

References

  • Abbas Edalat: Professor of Computer Science and Mathematics, Imperial College London, United Kingdom
  • Marco Wiering: Assistant Professor in Machine Learning, University of Groningen, The Netherlands