taohong08/2022-tnnls-fssc

Source Code for 'Fast Self-Supervised Clustering with Anchor Graph' (T-NNLS)

FSSC

Using the code, please cite: Wang J, Ma Z, Nie F, et al. Fast Self-Supervised Clustering with Anchor Graph[J]. IEEE Transactions on Neural Networks and Learning Systems, 33(9), pp. 4199-4212, 2022.

https://ieeexplore.ieee.org/document/9354504

The code explanation: The main function of the code: AnchorGEN.m and FSSC.m You can use demo.m to perform FSSC clustering on USPS and Letter data sets. If you have any questions, please connect [email protected]

Use of Main Function

To use function 'AnchorGEN.m' for constructing anchor graph, please follow the input/output format:

[B,M] = AnchorGEN(X,numAnchor,numNeighbor,generateAnchor)

Input: numAnchor: the number of anchors numNeighbor: the number of neighbors of anchor graph generateAnchor: option '1' means 'BKHK', '2' means 'kmeans++', '3' means 'kmeans', '4' means 'Random Selection'

Output: B: anchor graph M: anchor data matrix

Example for USPS(9298$\times$256) data set: [B,M] = AnchorGEN(X,9,20,1)

To use function 'FSSC.m' for self-supervised clustering, please follow the input/output format:

[result,labelnew,t,Rank,rp] = FSSC(X,B,M,label,alpha_u,alpha_l,isW)

Input: X: data matrix B: anchor graph matrix M: anchor data matrix label: ground truth (for compute clustering results) alpha_u: the coefficient of detecing novel class, default 0.99 alpha_l: the coefficient of changing primal label, default 0 isW: option '1' means 'compute W for sample similarity', option '0' means 'not compute W for reduce space complexity', default 1

Output: result: clustering results ACC NMI ARI labelnew: predicted label by FSSC t: running time Rank: the index of anchors from full samples rp: the index of c representative points from full samples

Example for USPS(9298$\times$256) data set: [result,labelnew,t,Rank,rp] = FSSC(X,B,M,label) result[ACC,NMI,ARI] = [0.7029 0.6636 0.6027]