Evolution-In-Materio: Solving Machine Learning Classification Problems Using Materials

详细信息    查看全文

• 作者：Maktuba Mohid (19)
  Julian Francis Miller (19)
  Simon L. Harding (20)
  Gunnar Tufte (20)
  Odd Rune Lykkeb酶 (21)
  Mark K. Massey (21)
  Michael C. Petty (19)
  
• 关键词：Evolutionary algorithm ; evolution ; in ; materio ; material computation ; evolvable hardware ; machine learning ; classification problem
• 刊名：Lecture Notes in Computer Science
• 出版年：2014
• 出版时间：2014
• 年：2014
• 卷：8672
• 期：1
• 页码：721-730
• 全文大小：373 KB
• 参考文献：1. Akbarzadeh, V., Sadeghian, A., dos Santos, M.: Derivation of relational fuzzy classification rules using evolutionary computation. In: IEEE Int. Conf. on Fuzzy Systems, pp. 1689鈥?693 (2008)
  2. Bache, K., Lichman, M.: UCI machine learning repository (2013), http://archive.ics.uci.edu/ml
  3. Broersma, H., Gomez, F., Miller, J.F., Petty, M., Tufte, G.: Nascence project: Nanoscale engineering for novel computation using evolution. International Journal of Unconventional Computing聽8(4), 313鈥?17 (2012)
  4. Harding, S., Miller, J.F.: Evolution in materio: A tone discriminator in liquid crystal. In: Proc. Congress on Evolutionary Computation 2004, vol.聽2, pp. 1800鈥?807 (2004)
  5. Harding, S., Miller, J.F.: Evolution in materio: A real time robot controller in liquid crystal. In: Proc. NASA/DoD Conference on Evolvable Hardware, pp. 229鈥?38 (2005)
  6. Harding, S.L., Miller, J.F.: Evolution in materio: Evolving logic gates in liquid crystal. Int. J. of Unconventional Computing聽3(4), 243鈥?57 (2007)
  7. Leitner, J., Harding, S., Forster, A., Schmidhuber, J.: Mars terrain image classification using cartesian genetic programming. In: 11th International Symposium on Artificial Intelligence, Robotics and Automation in Space, i-SAIRAS (2012)
  8. Lykkeb酶, O.R., Harding, S., Tufte, G., Miller, J.F.: Mecobo: A Hardware and Software Platform for In Materio Evolution. In: Ibarra, O.H., Kari, L., Kopecki, S. (eds.) UCNC 2014. LNCS, vol.聽8553, pp. 267鈥?79. Springer, Heidelberg (2014), http://dx.doi.org/10.1007/978-3-319-08123-6_22 CrossRef
  9. Miller, J.F. (ed.): Cartesian Genetic Programming. Springer (2011)
  10. Miller, J.F., Downing, K.: Evolution in materio: Looking beyond the silicon box. In: Stoica, A., Lohn, J., Katz, R., Keymeulen, D., Zebulum, R.S. (eds.) The 2002 NASA/DoD Conference on Evolvable Hardware, vol.聽7, pp. 167鈥?76. IEEE Computer Society (2002)
  11. Miller, J.F., Harding, S.L., Tufte, G.: Evolution-in-materio: evolving computation in materials. Evolutionary Intelligence聽7, 49鈥?7 (2014) CrossRef
  12. Thompson, A.: Hardware Evolution - Automatic Design of Electronic Circuits in Reconfigurable Hardware by Artificial Evolution. Springer (1998)
  13. V枚lk, K., Miller, J.F., Smith, S.L.: Multiple network CGP for the classification of mammograms. In: Giacobini, M., et al. (eds.) EvoWorkshops 2009. LNCS, vol.聽5484, pp. 405鈥?13. Springer, Heidelberg (2009) CrossRef
  
• 作者单位：Maktuba Mohid (19)
  Julian Francis Miller (19)
  Simon L. Harding (20)
  Gunnar Tufte (20)
  Odd Rune Lykkeb酶 (21)
  Mark K. Massey (21)
  Michael C. Petty (19)
  
  19. Department of Electronics, University of York, York, UK
  20. Department of Computer and Information Science, Norwegian University of Science and Technology, 7491, Trondheim, Norway
  21. School of Engineering and Computing Sciences and Centre for Molecular and Nanoscale Electronics, Durham University, UK
  
• ISSN：1611-3349

文摘

Evolution-in-materio (EIM) is a method that uses artificial evolution to exploit the properties of physical matter to solve computational problems without requiring a detailed understanding of such properties. EIM has so far been applied to very few computational problems. We show that using a purpose-built hardware platform called Mecobo, it is possible to evolve voltages and signals applied to physical materials to solve machine learning classification problems. This is the first time that EIM has been applied to such problems. We evaluate the approach on two standard datasets: Lenses and Iris. Comparing our technique with a well-known software-based evolutionary method indicates that EIM performs reasonably well. We suggest that EIM offers a promising new direction for evolutionary computation.