文件列表:

SOQcode\fun_BSOQ2.m (2556, 2013-07-20)
SOQcode\fun_CESOQ.m (1321, 2013-07-20)
SOQcode\fun_ClassIndentifier2Queues.m (452, 2013-07-20)
SOQcode\fun_CompareWithGroundTruth.m (2594, 2013-07-20)
SOQcode\fun_DCSOQ.m (1087, 2013-07-20)
SOQcode\fun_DifferenceBetweenQueues.m (782, 2013-07-20)
SOQcode\fun_DifferenceWithinQueues.m (1334, 2013-07-20)
SOQcode\fun_DisplayQueues.m (691, 2013-07-20)
SOQcode\fun_DisplayQueuesBeforeRandperm.m (581, 2013-07-20)
SOQcode\fun_InitializeQueues.m (857, 2013-07-20)
SOQcode\fun_LoadData.m (4781, 2013-07-21)
SOQcode\fun_PlotGreyW.m (786, 2013-07-20)
SOQcode\fun_QueuesResultSorted.m (465, 2013-07-20)
SOQcode\fun_ReOrganizeQueues.m (3197, 2013-07-20)
SOQcode\fun_SelfOrganizingQueueWithInitialQueues.m (2423, 2013-07-20)
SOQcode\Main.m (1999, 2013-07-20)
SOQcode (0, 2013-07-20)

Version 20121029 Written by Baohua Sun Email: bsun@ufl.edu This folder contains the matlab code for Self-Organizing-Queue (SOQ) Based Clustering. The detailed description of the algorithm is downloadable at: http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6334425&url=http%3A%2F%2Fieeexplore.ieee.org%2Fiel5%2F97%2F6323087%2F06334425.pdf%3Farnumber%3D6334425 -------------------------------------- Main.m is a demo to show the performance of MSSOQ on 2-Dimensional Gaussian Data with mean value at (-0.3,0),(0,0),(0.3,0),(0.6,0) and equal variance of 0.05. And the number of points generated in each cluster is 105,15,15,15 respectively. Run Main.m (press F5) and you will see MSSOQ can give perfect clustering result in this experiment. In Main.m, there is a random permutation process to simulate the input with random order. Thus, this random order is taken care of when we call "fun_CompareWithGroundTruth(Queues,Index,QueuesGroundTruth)" to evaluate the result compared with the given ground truth. If you do not need to simulate random order, you can simply comment "Index=randperm(length(W));" and uncomment "Index=1:(length(W));", and then you don't need to modify other codes. -------------------------------------- If you want to implement MSSOQ on other data sets, feed your input similarity matrix (i.e. W) and number of clusters(i.e. K) into fun_DCSOQ.m, and you will get the result stored in Queues, which is a K*1 cell array. Each cluster corresponds to a cell, and each cell contains all the IDs of the nodes belonging to the same cluster. It will be more convenient for you to implement MSSOQ on your data set if you follow the format given in 'fun_LoadData.m', where you can load your data in the second last line(i.e. below "otherwise"). Then, you can simply change "DataSourceSelection" in Main.m and run Main.m to get the result. -------------------------------------- The codes contained in this folder has registered a copyright. Copyright information is given below: Copyright 2013 University of Florida Research Foundation, Inc.. All rights reserved. Copyright Registration Number: TXu 1-848-969, effective from January 9, 2013

近期下载者：

相关文件：

评论：[我要评论] [举报此文件]

收藏者：