Graduate Research Assistant

University of Alberta · Edmonton, Canada · May 2024 – Present

  • Conduct research in reinforcement learning and active inference to model decision-making in non-stationary environments.
  • Compare active inference to human-like decision-making and classic RL agents in cooperative settings.
  • Implement a model-based active inference agent using predictive-coding principles and graphical modeling.
  • Apply variational Bayesian inference to the agent’s model to maintain beliefs about the environment.

Research Assistant

Isfahan University of Technology · Isfahan, Iran · Nov 2021 – Aug 2022

  • Extended the RACER classifier to handle continuous features using supervised discretization.
  • Evaluated CAIM, MDLP, CART, and ChiMerge methods for accuracy and interpretability.
  • Achieved a 6.6% accuracy gain and 10–20% fewer rules, surpassing Naive Bayes, LightGBM, and SVM.

Research Assistant

Isfahan University of Technology · Isfahan, Iran · Oct 2021 – Aug 2022

  • Built a dual-stream Mask R-CNN-based model for visual streams, unlike prior separate-stream models.
  • Simulated dorsal synaptic perturbations to model ASD/CVI distance-estimation deficits.
  • Demonstrated restoration of spatial perception through visual-therapy-style retraining.