{"id":5851,"date":"2011-10-10T13:17:19","date_gmt":"2011-10-10T10:17:19","guid":{"rendered":"http:\/\/hgpu.org\/?p=5851"},"modified":"2011-10-10T13:17:19","modified_gmt":"2011-10-10T10:17:19","slug":"data-parallel-construction-of-delta_n-nets-with-maximum-dispersion","status":"publish","type":"post","link":"https:\/\/hgpu.org\/?p=5851","title":{"rendered":"Data-Parallel Construction of delta_N-Nets with Maximum Dispersion"},"content":{"rendered":"<p>Linear nearest-neighbor search in high-dimensional data exposes high computational complexity. In order to minimize search complexity we employ optimal delta-nets of rank N, which consist of a small sub set of N vectors out of an initial code book E, yet approximate all En vectors of E by the least error of all possible selections of N vectors. By employing a distributed, data-parallel approach which can be computed efficiently on a cluster of GPUs, we observe speedup factors up to 215, compared to a sequential CPUbased solution. Despite the construction process for a suitable delta-net being of complexity O(EnN^2 &#8211; N^3), our algorithm makes computation of quasi-optimal delta-nets feasible with respect to our sample application of principal component analysis (PCA)-based feature detection.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Linear nearest-neighbor search in high-dimensional data exposes high computational complexity. In order to minimize search complexity we employ optimal delta-nets of rank N, which consist of a small sub set of N vectors out of an initial code book E, yet approximate all En vectors of E by the least error of all possible selections [&hellip;]<\/p>\n","protected":false},"author":351,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":false,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2}},"categories":[36,11,90,3],"tags":[1787,659,1782,349,20,953,379,1793,680,441,378],"class_list":["post-5851","post","type-post","status-publish","format-standard","hentry","category-algorithms","category-computer-science","category-opencl","category-paper","tag-algorithms","tag-computational-complexity","tag-computer-science","tag-nearest-neighbour","tag-nvidia","tag-nvidia-geforce-gtx-470","tag-nvidia-geforce-gtx-480","tag-opencl","tag-openmpi","tag-search","tag-tesla-c2050"],"views":1967,"jetpack_publicize_connections":[],"jetpack_featured_media_url":"","jetpack_sharing_enabled":true,"_links":{"self":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/5851","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/users\/351"}],"replies":[{"embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=5851"}],"version-history":[{"count":0,"href":"https:\/\/hgpu.org\/index.php?rest_route=\/wp\/v2\/posts\/5851\/revisions"}],"wp:attachment":[{"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=5851"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=5851"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/hgpu.org\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=5851"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}