Student Research Scientist @ RIACS/NASA Ames Research Center
Master of Science (M.S.), Information and Computer Science @
University of California, Irvine
I am a statistical machine learning professional with 16 years of experience (10 years post-Ph.D.) in modeling and understanding of large-scale noisy data in high-impact applications. My previous work spans applications of probabilistic approaches and Bayesian techniques to high-dimensional data understanding, with applications to modeling of atmospheric data, graph/networks, and more recently, to ranking by relevance for
I am a statistical machine learning professional with 16 years of experience (10 years post-Ph.D.) in modeling and understanding of large-scale noisy data in high-impact applications. My previous work spans applications of probabilistic approaches and Bayesian techniques to high-dimensional data understanding, with applications to modeling of atmospheric data, graph/networks, and more recently, to ranking by relevance for product search, and to financial and security applications.
My contact information, list of publications, and my full academic CV can be found at http://www.sergeykirshner.com.
Principal Member of Technical Staff @ My current responsibilities include:
- investigation and prototyping of machine learning approaches for the purpose of including them into the Skytree Infinity machine learning platform,
- development of IP (filed 3 provisional patents),
- fundamental research of machine learning techniques related to data preprocessing, feature extraction and selection, and model selection and estimation,
- leading and consulting on Skytree customers' projects. From November 2014 to Present (1 year 2 months) Senior Machine Learning Researcher @ Research of brain-inspired machine learning approaches for computer vision. From August 2014 to October 2014 (3 months) Software Development Engineer / Machine Learning Scientist @ As a part of a team, I was improving Amazon.com's product search engine. This included:
- Conceptual development of approaches for learning of product search relevance ranking functions from the massive catalog and customer log data;
- Implementation and deployment of the above approaches under strict search latency requirements (deployed 7 different ranking functions spanning in total about 7% of the Amazon's worldwide retail website product search by relevance by query volume volume);
- Development of the new and refinement of the existing features for product search relevance rankings;
- A/B hypothesis testing and analysis and interpretation of experimental results;
- Development of intellectual property and internal dissemination of machine learning expertise;
- Communication with business teams as a part of planning and decision making. From August 2013 to August 2014 (1 year 1 month) Assistant Professor @ As an Assistant Professor, I pursued original research ideas (conception, funding, development, publishing) in collaboration with other faculty and solo, trained graduate and undergraduate students, developed and taught graduate and undergraduate level courses, and served on departmental committees.
Most of my research revolved around the statistical machine learning approaches and software for modeling of high-dimensional data with strong emphasis on probabilistic models. A number of my ideas and software were applied to problems in atmospheric sciences and hydrology. (See publications.)
I supervised a small group (3-4) of graduate and undergraduate students. Our work was in part supported by the grants from the US National Science Foundation (IIS-0916686 and AGS-1025430) as well as internal Purdue funding.
My standard teaching load was 3 semester courses per year. I taught undergraduate and masters level probability (6 sessions), masters level computational statistics (3 sessions), and a graduate level probabilistic graphical models (3 sessions) courses. I also ran several seminar courses and reading groups. In addition to that, together with my colleagues, I co-organized a Machine Learning Summer School in 2011. This event attracted 130 students and professionals who for two weeks immersed themselves in lectures delivered by the world's machine learning experts. From August 2008 to August 2013 (5 years 1 month) Postdoctoral Fellow @ From June 2006 to July 2008 (2 years 2 months) Postdoctoral Researcher @ From April 2005 to May 2006 (1 year 2 months) Graduate Research Assistant @ From April 2001 to March 2005 (4 years) Student Research Scientist @ From June 2000 to August 2000 (3 months) Research Programmer @ From June 1998 to August 1999 (1 year 3 months) Summer Intern @ From June 1997 to August 1997 (3 months)
Doctor of Philosophy (Ph.D.), Information and Computer Science @ University of California, Irvine From 2001 to 2005 Master of Science (M.S.), Information and Computer Science @ University of California, Irvine From 1999 to 2001 Bachelor of Arts (B.A.), Mathematics and Computer Science @ University of California, Berkeley From 1995 to 1998 Sergey Kirshner is skilled in: Machine Learning, Statistics, Data Mining, Data Science, Artificial Intelligence, Algorithms, Computer Science, Time Series Analysis, Multivariate Statistics, Graphical Models, Hidden Markov Models, Copulas, Python, C, C++
Looking for a different
Get an email address for anyone on LinkedIn with the ContactOut Chrome extension