Optimal control, no! Than what? (...)
summary
Covers concrete algorithms and tradeoffs: simulated annealing finds near-global minima by checking ~360,000 samples in a 30‑dimensional test, Particle Swarm Optimization hit good approximations with ~5,000 tries, and differential evolution/genetic algorithms use population tuning (examples showed 300k evaluations). Shows runnable Java examples from GitHub (JSwarm, simple libraries) and contrasts them with heavy end‑to‑end deep RL (mentions DL4J / RL4J and DeepMind Atari/AlphaGo). Recommends simpler metaheuristics or a Python microservice before full deep RL.
AI-written summary · about
chapters (4)
related talks
30mEvolutionary Algorithms Perfecting the Art of "Good Enough"
54mAmazing Algorithms for Solving Problems in Software
07mSwarm Intelligence Algorithms - Session Interview With Frances Buontempo
01h 15mAn Introduction to Swarm Intelligence Algorithms
34mAlgorithms to live
24mIntelligent system optimizations
15mMulti-Agent Reinforcement Learning In Stochastic Games: From Alphago To Robust Control
01h 01mIntroduction to Practical AI