PPT-Kernel Learning with a Million Kernels

S VN Vishwanathan Purdue University Ashesh Jain IIT Delhi Manik Varma Microsoft Research India To appear SIGKDD 2012 The objective in kernel learning is to jointly learn both SVM and kernel parameters from training data. Solar . Radiative. Kernel. s. And Applications. Zhonghai Jin. Constantine . Loukachine. Bruce . Wielicki. Xu. Liu. SSAI, Inc. / NASA . Langley research . Center. July 6-9, 2010. Objective:. . Introduce the reflected solar spectral kernels, their spectral characteristics, and the potential applications to CLARREO . Also known as “meta-data”. April 2016. Comments in SPICE Kernels. 2. Comments, also called “meta-data,” are information that describe the context of kernel data, i.e. “data about data”. Comments are provided inside kernels as plain text (prose). April 2016. Porting Kernels. 2. Porting Issues - 1. Data formats vary across platforms, so data files created on platform “X” may not be usable on platform “Y.”. Binary. . formats. : different platforms use different bit patterns to represent numbers (and possibly characters).. kernels for finite-frequency signals: Applications in migration velocity updating and tomography. Xiao-Bi . Xie. University of California at Santa Cruz. Sanya. , China July 24-28, 2011. A brief introduction. April 2016. Agenda. Overview. Kernel architecture. Producing kernels. Using kernels. Introduction to Kernels. 2. Introduction to Kernels. 3. What is a SPICE “Kernel”. “Kernel” means file. “Kernel” means a file . Registration. Hw2. is out . Please start working on it as soon as possible. Come to sections with questions. On Thursday (TODAY) we will have two lectures:. Usual one, 12:30-11:45. An additional one, . 0.2 0.4 0.6 0.8 1.0 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 kernel(b) kernel(c) kernel(d) (a)blurredimage(b)no-blurredimage0.900.981.001.021.10 (5.35,3.37)(4.80,3.19)(4.71,3.22)(4.93,3.23)(5.03,3.22 S. ynthesis from . OpenCL. using Reconfiguration . C. ontexts. Authors:. James . Coole. , Greg . Stitt. University of Florida. Dept. . of Electrical . & . Computer Engineering and. NSF . CHREC. Gainesville, FL, USA. Specialization of Kernelization. Daniel Lokshtanov. We. Kernels. ∃. ¬. . Kernels. Why?. What’s Wrong with Kernels. (from a practitioners point of view). Only handles . NP-hard. problems.. Don’t combine well with . Machine Learning. March 25, 2010. Last Time. Basics of the Support Vector Machines. Review: Max . Margin. How can we pick which is best?. Maximize the size of the margin.. 3. Are these really . “equally valid”?. conjunctions . the learner is to learn. The number of . conjunctions. : . . log(|C. |) = . n. The elimination algorithm makes . n. . mistakes. Learn from . positive . examples; eliminate active literals. Machine Learning. March 25, 2010. Last Time. Recap of . the Support Vector Machines. Kernel Methods. Points that are . not. linearly separable in 2 dimension, might be linearly separable in 3. . Kernel Methods. :. Native & Convenient Heterogeneous. Computing for D. Outline. Introduction. Compiler. Libraries. Using . DCompute. (present and future). Future directions. State of Hardware. X86 . –. all compilers. Ke Wang. Sparse Correspondence Problems. Dense Correspondence Problems. Stereo. Motion. Motion vs. Stereo: Differences. Motion: . Uses velocity: consecutive frames must be close to get good approximate time derivative.