JDD 2018: Optimal control, no! Than what? (...) by Tomasz sikora

The general optimisation, as a search problem, is it only academic? No!

Life is being almost a one big pursuit of optimisation. Then why the development practice shows it is so challenging to adopt the mathematical thinking in the business terms? How to roll-out the concept to the business area, solve a real problem, and make your product owner a hero?

Well, that is your software craftsman developer duty, but I will drive you through a few optimisation techniques: starting from gradient descent as a skier on a slope, through local search to nature inspired optimisation, reviewing scale of search spaces, going via Simulated Annealing, Evolutionary Programming, Swarm Intelligence, Particle Swarm Optimisation, touching on observer problem and effective control challenges, to be able to finish on Reinforcement Learning and trying to answer questions why there is so much hype on the topic now, why it is called hedonistic search, and why not to begin with deep learning techniques (DL4J / RL4J) to solve any problem ;).