Dan Kushnir
Research Interests
- Artificial Intelligence
- Computational & Algorithmic Sciences
- Graph Theory
- Machine Learning
- Statistical & Data Sciences
Education
Ph.D. in Applied Mathematics, Thesis: Multiscale tools for data analysis, Advisors: Prof. Achi Brandt and Prof. Edriss Titi, Weizmann Institute of Science, Rehovot, Israel. 2004-2008
M.Sc. in Applied Mathematics, Thesis: A fast multi-scale clustering algorithm with application to cold and dark matter simulations, Advisor: Prof. Achi Brandt, Weizmann Institute of Science, Rehovot,Israel, 2001-2004
B.Sc. in Computer Science, The Hebrew University, Jerusalem, Israel, 1997-2001
Professional activities
Member of the Center for Discrete Mathematics and Computer Science (DIMACS) at Rutgers University
Organizer of the Applied Mathematics Seminar at Yale University
Member of the Society of Industrial and Applied Mathematics (SIAM)
Review Writer for Mathematical Reviews (MR) of the American Mathematical Society (AMS)
PhD thesis examiner
Reviewer in the following venues and journals:
Conferences: NeurIps, ICML, ICLR, AAAI, AISTATS, ISIT.
Journal of Machine Learning Research (JMLR)
IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
Applied and Computational Harmonic Analysis (ACHA)
SIAM Journal on Scientific Computing (SISC)
SIAM Journal on Matrix Analysis and Applications
Numerical Linear Algebra with applications
Journal of Global Optimization
IEEE Signal Processing Letters
IEEE Intelligent Systems
Computers and Mathematics with Applications
Bell Labs Technical Journal
Selected articles and publications
D. Kushnir, B. Mirabelli, Active Community Detection with Maximal Expected Model Change, AISTATS, 2020 (oral presentation)
D. Kushnir, S. Jalali, I. Saniee, Towards Clustering High-dimensional Gaussian Mixture Clouds in Linear Running Time. AISTATS 2019
V. Gurbani, D. Kushnir, V. Mendiratta, C. Phadke, E. Falk, R. State, Detecting and Predicting Outage in Mobile Networks with Log Data, ICC, 2017
D. Kushnir, M. Goldstein, Causality inference for Failures in NFV, INFOCOM workshops, 929-934 ,2016
A. Akyamac, C. Phadke, D. Kushnir, H. Uzunalioglu, Predicting Home Network Problems Using Diverse Data, Sarnoff Symposium, 1-6, 2015
D. Kushnir, Efficient Active Learning with Label-Adapted Kernels, ACM 20th conference on Knowledge Discovery and Data Mining (ACM SIGKDD 2014), NYC.
C. Phadke, H. Uzunalioglu, V. Mendiratta, D. Kushnir, D. Doran, Prediction of Subscriber Churn Using Social Network Analysis, Bell Labs Technical Journal, (BLTJ), 17(4), 63-75, 2013
R. Talmon, D. Kushnir, R. R. Coifman, Israel Cohen, Sharon Gannot, Parametrization of Linear Systems Using Diffusion Kernels, IEEE Transactions on Signal Processing, 60(3), (2012)
A. Haddad, D. Kushnir, R. R. Coifman, Filtering via a reference set, revised for Applied and Computational Harmonic Analysis (ACHA) (2011)
D. Kushnir, A. Haddad, R. R. Coifman, Anisotropic diffusion on sub-manifolds with application to
Earth structure classification, Applied and Computational Harmonic Analysis (ACHA), 32(2), 280–294 (2012)
D. Kushnir, V. Rokhlin, A highly accurate stiff ODEs solver, SIAM Journal on Scientific Computing
(SISC), 34(3), 1296–1315 (2012)
D. Kushnir, M. Galun, A. Brandt, Efficient multilevel eigensolvers with applications to data analysis tasks, IEEE Transaction on Pattern Analysis and Machine Intelligence (TPAMI), vol. 32(8), 1377–1391 (2010)
Y. Goldberg, A. Zakai, D. Kushnir, Y. Ritov, Manifold Learning: the price of normalization, Journal of Machine Learning Research (JMLR), vol. 9, 1909-1939 (2008)
D. Kushnir, M. Galun, A. Brandt, Fast multiscale clustering and manifold identification, special issue on similarity-based pattern recognition, Pattern Recognition, 39(10), 1876–1891 (2006)
D. Kushnir, J. Schumacher, A. Brandt, Geometry of intensive scalar dissipation events in turbulence, Physical Review Letters 97, 124502 (2006) (cover page)
J. Schumacher, D. Kushnir, A. Brandt, K. R. Sreenivasan, and H. Zilken, Statistics and geometry in high-Schmidt number scalar mixing, Proceedings of the iTi Conference in Turbulence, Progress in Turbulence II, Springer Proceedings in Physics (2005)
Patents
NC324696 ACTO: An Efficeint System For Data Annotation And Model Building
NC321942 Focused training of deep neural nets for classification of imagery data, filed, 2020
NC317813 On-the-fly Active Learning for Acceleration of Deep Learning Training for Video and Imagery Data Classification, filed, 2019
NC307291 Active deep learning with applications to image and data classification, 2018
10541880 Control of data reporting for a data analytics service using an active learning framework
NC103295 Active Learning Framework and Architecture for IoT Analytics Services, granted, 2018
820588-EP-EPA A Method for Identifying Causality Objects, filed
818893-US-NP Alarm Causality Templates for Network Function Virtualization
816508-US-NP Root Cause Analysis For Service Degradation In Computer Networks (US9424121)
818347-US-NP Optimized Radio Frequency Signal Strength Sampling Of A Broadcast Area For Device Localization, (US9571976)
812168US-NP System and Method for Generating Subscriber Churn Predictions, (US8804929)
814019US-NP Data Classifier Using Proximity Graphs, Edge Weights, And Propagation Labels (US20140317034)
814359US-NP Adaptive Polling of Information from a Device (US9032119)
810349US-NP Method and Apparatus for Deriving Composite Tie Metric for Edge Between Nodes of Telecommunication Call Graph, (US9159077)
815307-US-NP Classification Of Device Problems Of Customer Premises Devices, (US20150278823)