Name: Paul Templier
Type: User
Company: @Sureli
Bio: PhD in evolutionary neural policy search at ISAE-SUPAERO, Toulouse, France.
Interested in ML, RL, evolution, robotics and over-engineered python code.
Location: Toulouse, France
Blog: https://templierpaul.github.io/online-cv/
Paul Templier's Projects
WIP - Basis for a custom auto ML framework to help with data processing and ML models tests
An Evolutionary Computation framework
Cartesian Genetic Programming for Julia
CEM-RL with gym 0.26 support and wandb logging
Python tool to run experiments on remote hosts
2023 Systems Capture the Flag
Dopamine is a research framework for fast prototyping of reinforcement learning algorithms.
Evolutionary Algorithm bot for the Breezy Project competition (https://web.cs.dal.ca/~dota2/?page_id=353)
Deep Learning-based simulator of DOTA 2 for ML training purposes.
Introduction to Evolutionary Computation class
Julia implementations of Natural Evolutionary Strategies and CMA-ES, based on Cambrian
Evolutionary strategies applied to policy search in deep reinforcement learning
A geometric encoding for neural network evolution
Artificial Life in a 2D environment based on NeuroEvolution of Augmenting Topologies
Generic genetic algorithm basis to try ideas, with a TSP implementation for evaluation
JRPG-style game created with the open-source game design tool Godot (WIP)
Research project with Yiming Han at ISAE-Supaero with Dassault Aviation. Guided by Nicolas DROUGARD <[email protected]> and Caroline CHANEL <[email protected]>
Web scraper to get the most recent SoloQ results for League of Legends European Championship (LEC) players
Python server to schedule tasks on a machine using slurm-like commands.
My main Machine Learning class
Multi-Objective Evolutionary Algorithm
Multidimensional genetic programming for multiclass classification
Julia implementation of NEAT and HyperNEAT