HyperNEAT explained

Hypercube-based NEAT, or HyperNEAT,^[1] is a generative encoding that evolves artificial neural networks (ANNs) with the principles of the widely used NeuroEvolution of Augmented Topologies (NEAT) algorithm developed by Kenneth Stanley.^[2] It is a novel technique for evolving large-scale neural networks using the geometric regularities of the task domain. It uses Compositional Pattern Producing Networks ^[3] (CPPNs), which are used to generate the images for Picbreeder.org and shapes for EndlessForms.com . HyperNEAT has recently been extended to also evolve plastic ANNs ^[4] and to evolve the location of every neuron in the network.^[5]

Applications to date

• Multi-agent learning^[6]
• Checkers board evaluation^[7]
• Controlling Legged Robots^{[8] [9] [10] [11] [12] [13]} video
• Comparing Generative vs. Direct Encodings^{[14] [15] [16]}
• Investigating the Evolution of Modular Neural Networks^{[17] [18] [19]}
• Evolving Objects that can be 3D-printed^[20]
• Evolving the Neural Geometry and Plasticity of an ANN^[21]

External links

• HyperNEAT Users Page
• Ken Stanley's website
• "Evolutionary Complexity Research Group at UCF"
• NEAT Project Homepage
• PicBreeder.org
• EndlessForms.com
• BEACON Blog: What is neuroevolution?

Notes and References

1. Stanley. Kenneth O.. D'Ambrosio. David B.. Gauci. Jason. 2009-01-14. A Hypercube-Based Encoding for Evolving Large-Scale Neural Networks. Artificial Life. 15. 2. 185–212. 10.1162/artl.2009.15.2.15202. 19199382. 26390526. 1064-5462.
2. Stanley. Kenneth O.. Miikkulainen. Risto. 2002-06-01. Evolving Neural Networks through Augmenting Topologies. Evolutionary Computation. 10. 2. 99–127. 10.1162/106365602320169811. 1063-6560. 12180173. 10.1.1.638.3910. 498161.
3. Stanley. Kenneth O.. 2007-05-10. Compositional pattern producing networks: A novel abstraction of development. Genetic Programming and Evolvable Machines. en. 8. 2. 131–162. 10.1007/s10710-007-9028-8. 1389-2576. 10.1.1.643.8179. 2535195.
4. Book: From Animals to Animats 11. Risi. Sebastian. Stanley. Kenneth O.. Indirectly Encoding Neural Plasticity as a Pattern of Local Rules . 2010-08-25. Springer Berlin Heidelberg. 9783642151927. Doncieux. Stéphane. Lecture Notes in Computer Science. 6226 . 533–543. en. 10.1007/978-3-642-15193-4_50. Girard. Benoît. Guillot. Agnès. Hallam. John. Meyer. Jean-Arcady. Mouret. Jean-Baptiste. 10.1.1.365.5589.
5. Risi. Sebastian. Stanley. Kenneth O.. 2012-08-31. An Enhanced Hypercube-Based Encoding for Evolving the Placement, Density, and Connectivity of Neurons. Artificial Life. 18. 4. 331–363. 10.1162/ARTL_a_00071. 22938563. 3256786. 1064-5462. free.
6. Book: D'Ambrosio. David B.. Stanley. Kenneth O.. Proceedings of the 10th annual conference on Genetic and evolutionary computation . Generative encoding for multiagent learning . 2008-01-01. GECCO '08. New York, NY, USA. ACM. 819–826. 10.1145/1389095.1389256. 9781605581309. 11507017.
7. J. Gauci and K. O. Stanley, “A case study on the critical role of geometric regularity in machine learning,” in AAAI (D. Fox and C. P. Gomes, eds.), pp. 628–633, AAAI Press, 2008.
8. Book: Risi. Sebastian. Stanley. Kenneth O.. Proceedings of the 15th annual conference on Genetic and evolutionary computation . Confronting the challenge of learning a flexible neural controller for a diversity of morphologies . 2013-01-01. GECCO '13. New York, NY, USA. ACM. 255–262. 10.1145/2463372.2463397. 9781450319638. 10.1.1.465.5068. 10308013.
9. Book: Clune. J.. Beckmann. B. E.. Ofria. C.. Pennock. R. T.. 2009 IEEE Congress on Evolutionary Computation . Evolving coordinated quadruped gaits with the HyperNEAT generative encoding . 2009-05-01. 2764–2771. 10.1109/CEC.2009.4983289. 978-1-4244-2958-5. 10.1.1.409.3868. 17566887.
10. Book: Clune. Jeff. Ofria. Charles. Pennock. Robert T.. Proceedings of the 11th Annual conference on Genetic and evolutionary computation . The sensitivity of HyperNEAT to different geometric representations of a problem . 2009-01-01. GECCO '09. New York, NY, USA. ACM. 675–682. 10.1145/1569901.1569995. 9781605583259. 16054567.
11. Yosinski J, Clune J, Hidalgo D, Nguyen S, Cristobal Zagal J, Lipson H (2011) Evolving Robot Gaits in Hardware: the HyperNEAT Generative Encoding Vs. Parameter Optimization. Proceedings of the European Conference on Artificial Life. (pdf)
12. Lee S, Yosinski J, Glette K, Lipson H, Clune J (2013) Evolving gaits for physical robots with the HyperNEAT generative encoding: the benefits of simulation. Applications of Evolutionary Computing. Springer. pdf
13. Book: Applications of Evolutionary Computation. Lee. Suchan. Yosinski. Jason. Glette. Kyrre. Lipson. Hod. Clune. Jeff. Evolving Gaits for Physical Robots with the HyperNEAT Generative Encoding: The Benefits of Simulation . 2013-04-03. Springer Berlin Heidelberg. 9783642371912. Esparcia-Alcázar. Anna I.. Lecture Notes in Computer Science. 7835 . 540–549. en. 10.1007/978-3-642-37192-9_54. 10.1.1.364.8979.
14. Clune. J.. Stanley. K. O.. Pennock. R. T.. Ofria. C.. 2011-06-01. On the Performance of Indirect Encoding Across the Continuum of Regularity. IEEE Transactions on Evolutionary Computation. 15. 3. 346–367. 10.1109/TEVC.2010.2104157. 1089-778X. 10.1.1.375.6731. 3008628.
15. Book: Parallel Problem Solving from Nature – PPSN X. Clune. Jeff. Ofria. Charles. Pennock. Robert T.. How a Generative Encoding Fares as Problem-Regularity Decreases . 2008-09-13. Springer Berlin Heidelberg. 9783540876991. Rudolph. Günter. Lecture Notes in Computer Science. 5199 . 358–367. en. 10.1007/978-3-540-87700-4_36. Jansen. Thomas. Beume. Nicola. Lucas. Simon. Poloni. Carlo.
16. Book: Advances in Artificial Life. Darwin Meets von Neumann. Clune. Jeff. Beckmann. Benjamin E.. Pennock. Robert T.. Ofria. Charles. HybrID: A Hybridization of Indirect and Direct Encodings for Evolutionary Computation . 2009-09-13. Springer Berlin Heidelberg. 9783642213137. Kampis. George. Lecture Notes in Computer Science. 5778 . 134–141. en. 10.1007/978-3-642-21314-4_17. Karsai. István. Szathmáry. Eörs. 10.1.1.409.741.
17. Book: Clune. Jeff. Beckmann. Benjamin E.. McKinley. Philip K.. Ofria. Charles. Proceedings of the 12th annual conference on Genetic and evolutionary computation . Investigating whether hyperNEAT produces modular neural networks . 2010-01-01. GECCO '10. New York, NY, USA. ACM. 635–642. 10.1145/1830483.1830598. 9781450300728. 10.1.1.409.4870. 14826185.
18. Book: Suchorzewski. Marcin. Clune. Jeff. Proceedings of the 13th annual conference on Genetic and evolutionary computation . A novel generative encoding for evolving modular, regular and scalable networks . 2011-01-01. GECCO '11. New York, NY, USA. ACM. 1523–1530. 10.1145/2001576.2001781. 9781450305570. 10.1.1.453.5744. 2542736.
19. Book: Verbancsics. Phillip. Stanley. Kenneth O.. Proceedings of the 13th annual conference on Genetic and evolutionary computation . Constraining connectivity to encourage modularity in HyperNEAT . 2011-01-01. GECCO '11. New York, NY, USA. ACM. 1483–1490. 10.1145/2001576.2001776. 9781450305570. 10.1.1.379.1188. 1442181.
20. Clune. Jeff. Lipson. Hod. 2011-11-01. Evolving 3D Objects with a Generative Encoding Inspired by Developmental Biology. ACM SIGEVOlution. 5. 4. 2–12. 10.1145/2078245.2078246. 9566239. 1931-8499.
21. Book: Risi. S.. Stanley. K. O.. The 2012 International Joint Conference on Neural Networks (IJCNN) . A unified approach to evolving plasticity and neural geometry . 2012-06-01. 1–8. 10.1109/IJCNN.2012.6252826. 978-1-4673-1490-9. 10.1.1.467.8366. 14268194.