Difference between revisions of "Ryan Rifkin"
GerdIsenberg (talk | contribs) |
GerdIsenberg (talk | contribs) |
||
| Line 34: | Line 34: | ||
[[Category:Chess Programmer|Rifkin]] | [[Category:Chess Programmer|Rifkin]] | ||
[[Category:Researcher|Rifkin]] | [[Category:Researcher|Rifkin]] | ||
| + | [[Category:Quotes|Rifkin]] | ||
Latest revision as of 15:46, 7 December 2019
Ryan Michael Rifkin,
an American mathematician and computer scientist with expertise in machine learning, affiliated with Google Inc. [1], Mountain View, California and previously at Honda Research Institute, Boston, and the MIT Center for Biological and Computational Learning.
He holds a Ph.D. from MIT on machine learning in 2002 [2]. In the early 90s, along with primary author Bradley Kuszmaul, Ryan Rifkin co-authored the massive parallel chess program StarTech which competed the ACM 1993 running on a CM-5 [3] [4]. He optimized some move generation routines in assembly language, and cleaned up the parallel work-stealing code.
Quotes
from Bradley Kuszmaul's Ph.D. thesis [5]
The code for move generation and checking illegal moves, which takes a total of 3.5% of the cycles, was optimized in assembly language by Ryan Rifkin under the direction of Mark Bromley of Thinking Machines Corporation. Before Ryan worked on that code, the move generation and illegal move checking accounted for about 9% of all the cycles.
Selected Publications
1998 ...
- Massimiliano Pontil, Ryan Rifkin, Theodoros Evgeniou (1998). From Regression to Classification in Support Vector Machines. A.I. Memo 1649, MIT Artificial Intelligence Lab, CiteSeerX
- Ryan Rifkin (1998). The Static Stochastic Ground-Holding Problem. Master's thesis, MIT
2000 ...
- Ryan Rifkin (2002). Everything Old Is New Again: A Fresh Look at Historical Approaches to Machine Learning. Ph.D thesis, MIT, pdf
- Ryan Rifkin, Aldebaro Klautau (2004). In Defense of One-Vs-All Classification. Journal of Machine Learning Research, Vol. 5, pdf
- Ross A. Lippert, Ryan Rifkin (2006). Infinite-σ Limits For Tikhonov Regularization. Journal of Machine Learning Research, Vol. 7, pdf [7]
- Ryan Rifkin (2007). Introduction to Fenchel Duality for Machine Learning. slides [8]
- Ryan Rifkin (2007). Fenchel Duality for Machine Learning, Part II: A Bit More Theory. slides
- Ryan Rifkin, Ross A. Lippert (2007). Value Regularization and Fenchel Duality. Journal of Machine Learning Research
- Ryan Rifkin, Ross A. Lippert (2007). Notes on Regularized Least-Squares. Technical Report, pdf
- Ryan Rifkin (2007). Regularized Least Squares. MIT Course, slides as pdf
- Ryan Rifkin (2009). Multiclass Classification. MIT Course, pdf
External Links
References
- ↑ Ryan M. Rifkin | Google Inc., Mountain View | Google
- ↑ Ryan M. Rifkin (2002). Everything Old Is New Again: A Fresh Look at Historical Approaches to Machine Learning. Ph. D thesis, MIT
- ↑ The 23rd ACM International Computer Chess Championship from The Computer History Museum, pdf
- ↑ Re: Hash tables----Clash!!!-What happens next? by Albert Gower, rec.games.chess, March 19, 1994
- ↑ Bradley C. Kuszmaul (1994). Synchronized MIMD Computing. Ph. D. thesis, Department of Electrical Engineering and Computer Science, MIT, pdf, pp. 130, 6.6 How Time is Spent in StarTech
- ↑ dblp: Ryan M. Rifkin
- ↑ Tikhonov regularization from Wikipedia named for Andrey N. Tikhonov
- ↑ Fenchel's duality theorem from Wikipedia named after Werner Fenchel
