Filters








9,239 Hits in 8.1 sec

DigitHist

Michael Shekelyan, Anton Dignös, Johann Gamper
2017 Proceedings of the VLDB Endowment  
We propose DigitHist, a histogram summary for selectivity estimation on multi-dimensional data with tight error bounds.  ...  By combining multi-dimensional and one-dimensional histograms along regular grids of different resolutions, DigitHist provides an accurate and reliable histogram approach for multi-dimensional data.  ...  In comparison, the u-error optimizes for tight bounds and also works well for more sophisticated spread assumptions.  ... 
doi:10.14778/3137628.3137658 fatcat:grfijk44wfb2bmeqchhbiitzie

A Learning Framework for Self-Tuning Histograms [article]

Raajay Viswanathan, Prateek Jain, Srivatsan Laxman, Arvind Arasu
2011 arXiv   pre-print
In particular, we show that SpHist obtains up to 50% less error than ISOMER on real-world multi-dimensional datasets.  ...  Specifically, we use query feedback from a workload as training data to estimate a histogram with a small memory footprint that minimizes the expected error on future queries.  ...  Multi-dimensional Histograms In this section, we empirically compare our EquiHist and SpHist methods with ISOMER for learning multi-dimensional histograms.  ... 
arXiv:1111.7295v2 fatcat:hte5dhg5dvh2thmxt7tysogzke

Transductive inference using multiple experts for brushwork annotation in paintings domain

Marchenko Yelizaveta, Chua Tat-Seng, Jain Ramesh
2006 Proceedings of the 14th annual ACM international conference on Multimedia - MULTIMEDIA '06  
In this paper, we develop a serial multi-expert framework for explicit annotation of paintings with brushwork classes.  ...  The selected features are utilized to generate several models to annotate the unlabelled patterns. The experts select the best performing model based on Vapnik combined bound.  ...  To improve the annotation accuracy in tasks with high-dimensional feature space, multi-expert frameworks have been proposed [12, 13] .  ... 
doi:10.1145/1180639.1180684 dblp:conf/mm/YelizavetaCJ06 fatcat:vytyetbfhrgklb5ivtprumt4k4

Improved selectivity estimation by combining knowledge from sampling and synopses

Magnus Müller, Guido Moerkotte, Oliver Kolb
2018 Proceedings of the VLDB Endowment  
In this paper, we present a novel approach to combine knowledge from synopses and sampling for the purpose of selectivity estimation for conjunctive queries.  ...  ., histograms, and, in addition, provide sampling facilities.  ...  For multi-dimensional synopses, we recommend histograms with tight bounds, e.g., [29] .  ... 
doi:10.14778/3213880.3213882 fatcat:fq6mtn5sybccfoobd4qiyuk2fi

Exploiting duality in summarization with deterministic guarantees

Panagiotis Karras, Dimitris Sacharidis, Nikos Mamoulis
2007 Proceedings of the 13th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD '07  
Histograms and several hierarchical techniques have been proposed for this problem.  ...  Compared to the state-of-the-art, our histogram construction algorithm reduces time complexity by (at least) a B log 2 n log * factor and our hierarchical synopsis algorithm reduces the complexity by (  ...  Error-bounded Histogram Construction We formulate the L w ∞ -bounded histogram construction problem: Problem 1 Given a data vector D and an L w ∞ -error bound , construct a histogram H of D with the minimum  ... 
doi:10.1145/1281192.1281235 dblp:conf/kdd/KarrasMS07 fatcat:fajb5mnqwrfqtiml4mgxa7gdle

Towards Symbolic Time Series Representation Improved by Kernel Density Estimators [chapter]

Matej Kloska, Viera Rozinajova
2021 Lecture Notes in Computer Science  
However, the disadvantage of this method is, that it works reliably only for time series with Gaussian-like distribution.  ...  Our goal was to optimally cover the information space by means of sufficient alphabet utilization; and to satisfy lower bounding criterion as tight as possible.  ...  TAILOR, a project funded by Horizon 2020 research and innovation programme under GA no 952215 and "Knowledge-based Approach to Intelligent Big Data Analysis" -Slovak Research and Development Agency under the  ... 
doi:10.1007/978-3-662-64553-6_2 fatcat:wfsspv2xgnhqrcnoawgxcp77ju

Synopses for Massive Data: Samples, Histograms, Wavelets, Sketches

Graham Cormode
2011 Foundations and Trends in Databases  
Extending from one-dimensional to multi-dimensional data, as with histograms, provides more definitional challenges.  ...  Methods for Approximate Query Processing (AQP) are essential for dealing with massive data.  ...  Acknowledgments The work of Minos Garofalakis was partially supported by the European Commission under FP7-FET Open (Future and Emerging Technologies) ICT-2009.8.0 grant no. 255957 (LIFT).  ... 
doi:10.1561/1900000004 fatcat:wk7razxkmzcv7fzczftlohblwa

Misfire Detection in a Spark Ignition Engine using Support Vector Machines

Babu S Devasenapati., K.I Ramachandran., V Sugumaran.
2010 International Journal of Computer Applications  
A comparative performance analysis on classification accuracy of SVM when using statistical and histogram features for misfire detection in a spark ignition engine is presented.  ...  To maintain optimum performance throughout the service life of an engine and to exercise a tight control over emissions, misfire detection is a vital activity.  ...  Dimensionality reduction Dimensionality reduction is the process of reducing the number of input features that are required for classification, done with the main intention of reducing the computational  ... 
doi:10.5120/917-1295 fatcat:m55y7ldaqzbhfi32lcnkxvtlye

SciBORQ: Scientific data management with Bounds On Runtime and Quality

Lefteris Sidirourgos, Martin L. Kersten, Peter A. Boncz
2011 Conference on Innovative Data Systems Research  
The ultimate goal is a complete system for scientific data exploration and discovery, capable of producing quality answers with strict error bounds in pre-defined time frames.  ...  An impression is selected such that the statistical error of a query answer remains low, while the result can be computed within strict time bounds.  ...  time and tight error bounds.  ... 
dblp:conf/cidr/SidirourgosKB11 fatcat:k7yis2qahbdvlk7gdpxtbxx66a

Mining compositional features for boosting

Junsong Yuan, Jiebo Luo, Ying Wu
2008 2008 IEEE Conference on Computer Vision and Pattern Recognition  
abilities, as well as bounded training error.  ...  These weak classifiers are further combined through a multi-class AdaBoost method for final multi-class classification.  ...  Wei Hao for visual feature extraction, Shengyang Dai and Liangliang Cao for helpful discussions. This work was supported in part by National Science Foundation Grants IIS-0347877 and IIS-0308222.  ... 
doi:10.1109/cvpr.2008.4587347 dblp:conf/cvpr/YuanLW08 fatcat:gsonfwjf7vgibmsgwlylkw4c7y

Secrets of GrabCut and Kernel K-Means

Meng Tang, Ismail Ben Ayed, Dmitrii Marin, Yuri Boykov
2015 2015 IEEE International Conference on Computer Vision (ICCV)  
Our bound formulation for kernel K-means allows to combine general pair-wise feature clustering methods with image grid regularization using graph cuts, similarly to standard color model fitting techniques  ...  We propose an alternative approach to color clustering using kernel K-means energy with wellknown properties such as non-linear separation and scalability to higher-dimensional feature spaces.  ...  First, we report results on the GrabCut database (50 images) using the bounding boxes provided in [20] . For each image the error is the percentage of mis-labeled pixels.  ... 
doi:10.1109/iccv.2015.182 dblp:conf/iccv/TangAMB15 fatcat:vj4ne2rmhjg6xpzaw3uwuo2gte

High dimensional nearest neighbor searching

Hakan Ferhatosmanoglu, Ertem Tuncel, Divyakant Agrawal, Amr El Abbadi
2006 Information Systems  
The techniques proposed in this paper are effective for real-life data sets, which are typically non-uniform, and they are scalable with respect to both dimensionality and size of the data set. r  ...  The evaluation establishes the superiority of the proposed techniques over the existing techniques for high dimensional similarity searching.  ...  Acknowledgements This material was prepared with the support of the US Department of Energy (DOE) Award No. DE-FG02-03ER25573.  ... 
doi:10.1016/j.is.2005.01.001 fatcat:z2kfzeqwfnekfpmt6fucje6w6a

Sample distribution shadow maps

Andrew Lauritzen, Marco Salvi, Aaron Lefohn
2011 Symposium on Interactive 3D Graphics and Games on - I3D '11  
Furthermore, the tighter SDSM shadow frusta enable more aggressive culling causing the algorithm to run faster than scene-independent schemes for complex geometry, such as this scene from Left for Dead  ...  Even with partitions tweaked to a specific camera view, PSSMs still produce lower quality shadows than SDSMs, which are fully automatic.  ...  Acknowledgements We would like to thank Jason Mitchell and Wade Schin from Valve Corporation for providing the Left 4 Dead 2 scene, and Jeffery Williams for the Tower scene.  ... 
doi:10.1145/1944745.1944761 dblp:conf/si3d/LauritzenSL11 fatcat:iymbpgax3fbhlmncjzzilwoitm

Obtaining tight bounds on higher-order interferences with a 5-path interferometer

Thomas Kauten, Robert Keil, Thomas Kaufmann, Benedikt Pressl, Časlav Brukner, Gregor Weihs
2017 New Journal of Physics  
Our results rule out the existence of higher order interference terms to an extent which is more than four orders of magnitude smaller than the expected pairwise interference, refining previous bounds  ...  Here, we perform such a test in optical multi-path interferometers.  ...  As is the case for any such null-test experiment, the tightness of the bound and, thereby, the strength of any conclusions to be drawn about the foundations of the theory depend on the measurement uncertainties  ... 
doi:10.1088/1367-2630/aa5d98 fatcat:immukv5mhnegrdidieeeo5nkxq

A data-adaptive and dynamic segmentation index for whole matching on time series

Yang Wang, Peng Wang, Jian Pei, Wei Wang, Sheng Huang
2013 Proceedings of the VLDB Endowment  
In addition to savings in space and time, our new index can provide tight upper and lower bounds on distances between time series.  ...  By dynamic segmentations adaptive to data, we can reduce dimensionality further, in this example, from 4 to 3 for S1 and S2, and to 2 for S3 and S4,.  ...  Lower Bound Tightness We tested the tightness of the proposed lower bound estimation approach.  ... 
doi:10.14778/2536206.2536208 fatcat:awj2kiq7tzh5vngxy6g5hvz7se
« Previous Showing results 1 — 15 out of 9,239 results