Piotr Wikieł - Do lazybones push the world forward?
0 likes63 views
This document discusses automation and its potential impacts, including a universal basic income. It then covers topics in soft computing like fuzzy logic, machine learning, and evolutionary computation techniques. Specifically, it explores evolutionary algorithms such as genetic algorithms and genetic programming. It describes how genetic programming works, including the function and terminal sets, mutation, crossover, and fitness functions used to evolve programs. Finally, it discusses applications of genetic programming and provides additional resources.
![Architecture-altering
operations
(if (< x 3)
(+ (* 2 x) 3)
(* (- x 2) x))
(defn foo1 [x]
(+ (* 2 x) 3))
(if (< x 3)
(foo1 x)
(* (- x 2) x))](https://image.slidesharecdn.com/meetit-en-161208193542/85/Piotr-Wikiel-Do-lazybones-push-the-world-forward-11-320.jpg)
Piotr Wikieł - Do lazybones push the world forward?
- 1. Do lazybones push the world forward? About programming automation Piotr Wikieł, MeetIT Toruń, 2016 @pwikiel, piotr.wikiel@gmail
- 2. „There's a pretty good chance we end up with a universal basic income, or something like that, due to automation” – Elon Musk
- 4. Exponential growth of computing. 20th to 21st centuries. CC-BY-1.0 Coutesy of Ray Kurzweil and Kurzweil Technologies, Inc.
- 5. Soft computing • Fuzzy Logic • Probabilistic Logic • Machine Learning • Evolutionary Computation
- 6. Evolutionary algorithms • Genetic algorithms • Genetic programming • Evolutionary programming • Neuroevolution • …
- 8. Function Set and Terminal Set
- 9. Mutation (if (< x 3) (+ (* 2 x) 3) (* (- x 2) x)) (if (< x 3) (+ (* 4 (/ x 2)) 3) (* (- x 2) x))
- 10. Crossover (if (< x 3) (+ (* 2 x) 3) (* (- x 2) x)) (if (< x 3) (+ (* 4 (/ x 2)) 3) (/ (* x Math/PI) 2)) (if (< x 3) (+ (* 2 x) 3) (/ (* x Math/PI) 2))
- 11. Architecture-altering operations (if (< x 3) (+ (* 2 x) 3) (* (- x 2) x)) (defn foo1 [x] (+ (* 2 x) 3)) (if (< x 3) (foo1 x) (* (- x 2) x))
- 12. Fitness function • Answers the question „How good is program in doing its work?” • May contain some penalties, e.g. for the size of the program • Lexicase selection
- 13. Search space b2 4 4ac WAT?
- 14. Push&PushGP • Language for creating evolutionary systems • Based on stacks: • one stack for every type of data • one stack for code • one stack for executed program • Based on operations on stacks’ elements.
- 15. (run-push ’(2 2 integer_add) (make-push-state)) :integer (4) (run-push ’(2 2 3 integer_add integer_mult) (make-push-state)) :integer (10)
- 16. Applications of genetic programming • Computations • Building of complex systems • Life modeling and simulations
- 17. • https://github.com/lspector • http://www.genetic-programming.com/ • http://faculty.hampshire.edu/lspector/push.html