Filters








271 Hits in 1.6 sec

Implementing FastMap on the GPU: Considerations on General-Purpose Computation on Graphics Hardware [article]

G. Reina, T.Ertl
2005 Computer Graphics and Visual Computing  
Such an approach can be considered a heuristic in itself, but has the advantage of being based on visual feedback, therefore allowing for iterative improvement of the results.  ...  As an example, we ported the dimensionality reduction algorithm FastMap to fragment programs and thus accelerated it by orders of magnitude, allowing for interactive tweaking and evaluating of the algorithm  ...  Acknowledgements We would like to thank Thomas Klein for his additional bandwidth benchmarking results and fruitful discussion.  ... 
doi:10.2312/localchapterevents/tpcg/tpcguk05/051-058 dblp:conf/tpcg/ReinaE05 fatcat:b2sraao24fcdjkwlmvljxcyzze

Sort-Merge Feature Selection for Video Data [chapter]

Yan Liu, John R. Kender
2003 Proceedings of the 2003 SIAM International Conference on Data Mining  
As the wrapper model, this paper exploits a novel combination of Fastmap for dimensionality reduction and Mahalanobis distance for likelihood determination.  ...  A novel algorithm called Basic Sort-Merge Tree (BSMT) is proposed to choose a very small subset of features for video classification in linear time in the number of features.  ...  For hand selection, 30 macro-blocks selected by hand, based on the intuition of the researchers. Figure 2 is a grand summary.  ... 
doi:10.1137/1.9781611972733.38 dblp:conf/sdm/LiuK03 fatcat:rdky7y24knc2zdvf5jq2avzram

Fast Video Retrieval under Sparse Training Data [chapter]

Yan Liu, John R. Kender
2003 Lecture Notes in Computer Science  
A novel combination of Fastmap for dimensionality reduction and Mahalanobis distance for likelihood determination is used as the induction algorithm.  ...  It combines the virtues of a wrapper model approach for better accuracy with those of a filter method approach for incrementally deriving the appropriate features quickly.  ...  It relies on the three algorithms working together in linear time of the features: Fastmap for dimensionality reduction, Mahalanobis distance for classification likelihood, and a sort-complement-merge  ... 
doi:10.1007/3-540-45113-7_40 fatcat:orpekhh2qnd4jf3tae62mzk3ja

Matrix eQTL: ultra fast eQTL analysis via large matrix operations

A. A. Shabalin
2012 Bioinformatics  
For modern datasets, eQTL analysis is a computationally intensive task as it involves testing for association of billions of transcript-SNP pair.  ...  As more transcripts and SNPs get interrogated over a growing number of samples, the demand for faster tools for eQTL analysis grows stronger.  ...  FastMap is a fast association mapping tool written in Java. It uses the discrete nature of genotype data to speed up calculations. FastMap is optimized for permutation based testing.  ... 
doi:10.1093/bioinformatics/bts163 pmid:22492648 pmcid:PMC3348564 fatcat:v7tda7drfnffnabjhxhep7fzcq

FastMapSVM: Classifying Complex Objects Using the FastMap Algorithm and Support-Vector Machines [article]

Malcolm C. A. White, Kushal Sharma, Ang Li, T. K. Satish Kumar, Nori Nakata
2022 arXiv   pre-print
FastMap is an efficient linear-time algorithm that maps complex objects to points in a Euclidean space while preserving pairwise domain-specific distances between them.  ...  However, they generally demand large amounts of time and data for model training; and their learned models can sometimes be difficult to interpret.  ...  MW and TKSK conceived the general concept of combining FastMap with SVMs and kernel methods for classification of complex objects, independent of [4] .  ... 
arXiv:2204.05112v3 fatcat:aekqmh4he5h63p43fobw44lhdm

Matrix eQTL: Ultra fast eQTL analysis via large matrix operations [article]

Andrey A. Shabalin
2011 arXiv   pre-print
We present a new method for fast eQTL analysis via linear models, called Matrix eQTL. Matrix eQTL can model and test for association using both linear regression and ANOVA models.  ...  A modern eQTL dataset contains millions of SNPs and thousands of transcripts measured for hundreds of samples.  ...  FastMap is optimized for permutation based testing. Note that FastMap is the only tool out of the five that has a graphical user interface.  ... 
arXiv:1105.5764v1 fatcat:fzis5rzozze3texdt36w4abpky

Optimizing Memory-mapped I/O for Fast Storage Devices

Anastasios Papagiannis, Giorgos Xanthakis, Giorgos Saloustros, Manolis Marazakis, Angelos Bilas
2020 USENIX Annual Technical Conference  
Memory-mapped I/O provides several potential advantages over explicit read/write I/O, especially for low latency devices: (1) It does not require a system call, (2) it incurs almost zero overhead for data  ...  We also show that FastMap is able to saturate state-of-the-art fast storage devices when used by a large number of cores, where Linux mmap fails to scale.  ...  Finally, we thank the anonymous reviewers for their insightful comments and our shepherd Sudarsun Kannan for his help with preparing the final version of the paper.  ... 
dblp:conf/usenix/PapagiannisXSMB20 fatcat:uyxuqvupezdt7azph66d3yxi7m

Learning Binary Codes For Efficient Large-Scale Music Similarity Search

Jan Schlüter
2013 Zenodo  
ACKNOWLEDGEMENTS The author would like to thank Mohammad Norouzi for publishing his HDML implementation, and Maarten Grachten for fruitful discussions.  ...  The Austrian Research Institute for Artificial Intelligence is supported by the Austrian Federal Ministry for Transport, Innovation, and Technology.  ...  FastMap serves as a prefilter like PCA, but supports non-vector models as it is purely distance-based.  ... 
doi:10.5281/zenodo.1416507 fatcat:7dytbacywjagbdgnypjeyeyjwe

Multidimensional Scaling Localization Algorithm in Wireless Sensor Networks

Zhang Dongyang, Bai Xue, Liu Pengnan
2014 Sensors & Transducers  
presents a fast multidimensional scaling location algorithm.  ...  By positioning algorithm for fast multidimensional scaling, fast mapping initialization, fast mapping and coordinate transform can get schematic coordinates of node, coordinates Initialize of MDS algorithm  ...  The triangulation algorithm based on propagation model is mostly used.  ... 
doaj:205f3da055754c2f951ea0a9ed0134f1 fatcat:f4dp3x2u7nbzpcwxpuhrdzf3ma

Application-Independent Localization Based on 3D Simulation Technology

Bjoern Sondermann, Markus Emde, Juergen Rossmann
2015 International Journal of Materials Mechanics and Manufacturing  
This testbed approach allows for the integration of real and simulated sensors in virtual and real testbeds for a smooth transition between simulation and real world tests.  ...  A generalization of the concept led to a highly modular framework that is adaptable to a multitude of new domains.  ...  The localization algorithm is based on a particle filter algorithm as described in [3] .  ... 
doi:10.7763/ijmmm.2016.v4.248 fatcat:jidkhp5k6jehpn5daaogjvu22y

Hierarchical Convex NMF for Clustering Massive Data

Kristian Kersting, Mirwaes Wahabzada, Christian Thurau, Christian Bauckhage
2010 Journal of machine learning research  
In this paper, we present a hierarchical CH-NMF that automatically adapts to internal structures of a dataset, hence it yields meaningful and interpretable clusters for non-convex datasets.  ...  We present an extension of convex-hull non-negative matrix factorization (CH-NMF) which was recently proposed as a large scale variant of convex non-negative matrix factorization or Archetypal Analysis  ...  Freeman for making the tiny images freely available and Salah Zayakh for preparing the DBLP dataset. Mirwaes Wahabzada and Kristian Kersting were supported by the Fraunhofer ATTRACT Fellowship STREAM.  ... 
dblp:journals/jmlr/KerstingWTB10 fatcat:uu3mbtdzovheznw6q3njs7na3e

Convex NMF on Non-Convex Massiv Data

Kristian Kersting, Mirwaes Wahabzada, Christian Thurau, Christian Bauckhage
2010 Lernen, Wissen, Daten, Analysen  
In this paper, we present a hierarchical CH-NMF that automatically adapts to internal structures of a dataset, hence it yields meaningful and interpretable clusters for non-convex datasets.  ...  We present an extension of convex-hull nonnegative matrix factorization (CH-NMF) which was recently proposed as a large scale variant of convex non-negative matrix factorization (C-NMF) or Archetypal Analysis  ...  Acknowledgment The authors would like to thank A. Torralba, R. Fergus, and W. T. Freeman for making the tiny images freely available and S. Zayakh for preparing the DBLP dataset. M. Wahabzada and K.  ... 
dblp:conf/lwa/KerstingWTB10 fatcat:lxanrmgfcnc5nncbamvjp6ygzm

Properties of embedding methods for similarity searching in metric spaces

G.R. Hjaltason, H. Samet
2003 IEEE Transactions on Pattern Analysis and Machine Intelligence  
SparseMap is a variant of Lipschitz embeddings, while FastMap and MetricMap are inspired by dimension reduction methods for Euclidean spaces (using KLT or the related PCA and SVD).  ...  Thus, queries can be performed (for the most part) on the embedded objects.  ...  Modifying the nonlinear optimization model to account for the heuristic, we obtained an expansion of up to seven for two iterations of FastMap, and 14 for three iterations.  ... 
doi:10.1109/tpami.2003.1195989 fatcat:h7lcrevkjzgprisnl4r3ngthdy

Anonymous Biometric Access Control

Shuiming Ye, Ying Luo, Jian Zhao, Sen-ChingS Cheung
2009 EURASIP Journal on Information Security  
the k candidates followed by a HE-based matching protocol applied only on these candidates.  ...  The ABAC system uses novel Homomorphic Encryption (HE) based protocols to verify membership of a user without knowing his/her true identity.  ...  Based on the security model introduced in Section 3, this protocol is secure.  ... 
doi:10.1186/1687-417x-2009-865259 fatcat:5drq3zervffh5n2zdtgytcr5mu

Anonymous Biometric Access Control

Shuiming Ye, Ying Luo, Jian Zhao, Sen-Ching S. Cheung
2009 EURASIP Journal on Information Security  
the k candidates followed by a HE-based matching protocol applied only on these candidates.  ...  The ABAC system uses novel Homomorphic Encryption (HE) based protocols to verify membership of a user without knowing his/her true identity.  ...  Based on the security model introduced in Section 3, this protocol is secure.  ... 
doi:10.1155/2009/865259 fatcat:bqv7abbx6bgjto6nsygz756hm4
« Previous Showing results 1 — 15 out of 271 results