Toolkits for feature extraction and characterization of network data
Date
2015-03-29
Authors
Bandara, Vidarshana W.
Journal Title
Journal ISSN
Volume Title
Abstract
Description
Zip file Data 1: GUI for Robust PCA recoverability experiments. The GUI provides the following functionalities: - Evaluate sufficient conditions for recovery over a selected range of ranks and sparsities, size, low-rank and sparse matrix types; - Recoverable region for a selected range fractional sparsities, size, low-rank and sparse matrix types; - Input - output mapping between fractional-ranks fractional-sparsities; - Recovery error of the low-rank component; - Recovery error of the sparse component.
Zip file Data 2: GUI for Robust Principal Component Analysis. The GUI performs Robust PCA on: - Synthesized low-rank and sparse matrix additions; - Data from external experiments.
Zip file Data 3: Random matrix generator. The software synthesizes random realizations of low-rank and sparse matrices of specified types. Synopsis: Input arguments: LS typ n frs reps fname LS: specify either low-rank or sparse. Use L or l to indicate low-rank matrices, and S or s to indicate sparse matrices. typ: indicates the type of the matrices. Types available for low-rank are: 1. First order Gaussian; 2. Second order Gaussian; 3. Wishart; 4. First order Vandermonde; 5. Second order Vandermonde; Type available for sparse matrices are: 1. Fixed; 2. Uniform; 3. Gaussian; n: size of the matrices. frs: fractional rank or fractional sparsity of the matrix. reps: number of realizations of the specified matrix. fname: base file name of the matrices. reps many files will be produced each containing a comma-delimited realization of the specified matrix. E.g.: If fname is testfile and reps is 3, then three output files with names testfile.1, testfile.2, and testfile.3 will be produced. Examples DOS: genRandMats-1.0.exe L 1 20 0.1 3 testfile Linux: ./run_genRandMatix_v01a.sh $MCR_DIRECTORY L 1 20 0.1 3 testfile.
Zip file Data 4: MATLAB® Toolkit for network traffic anomaly analysis. The toolkit performs: - De-trending and thresholding for anomaly detection; - Graph wavelet based summarizing and anomaly tracing; - Distribution fitting to spatial and temporal parameters; - Simulator/Emulator to regenerate statistically similar anomalies The toolkit is developed for Internet2 dataset, but customizable for other datasets.
Zip file Data 5: Source codes for Robust PCA experiments. MATLAB codes for Robust PCA related experiments. The codes generate random matrices and perform Robust PCA. The codes cover a range of decomposition and recovery experiments for RPCA.
Zip files 1-3 contain executable files; zip files 4-5 contain data and ReadMe files.
Department of Electrical and Computer Engineering
Zip file Data 2: GUI for Robust Principal Component Analysis. The GUI performs Robust PCA on: - Synthesized low-rank and sparse matrix additions; - Data from external experiments.
Zip file Data 3: Random matrix generator. The software synthesizes random realizations of low-rank and sparse matrices of specified types. Synopsis: Input arguments: LS typ n frs reps fname LS: specify either low-rank or sparse. Use L or l to indicate low-rank matrices, and S or s to indicate sparse matrices. typ: indicates the type of the matrices. Types available for low-rank are: 1. First order Gaussian; 2. Second order Gaussian; 3. Wishart; 4. First order Vandermonde; 5. Second order Vandermonde; Type available for sparse matrices are: 1. Fixed; 2. Uniform; 3. Gaussian; n: size of the matrices. frs: fractional rank or fractional sparsity of the matrix. reps: number of realizations of the specified matrix. fname: base file name of the matrices. reps many files will be produced each containing a comma-delimited realization of the specified matrix. E.g.: If fname is testfile and reps is 3, then three output files with names testfile.1, testfile.2, and testfile.3 will be produced. Examples DOS: genRandMats-1.0.exe L 1 20 0.1 3 testfile Linux: ./run_genRandMatix_v01a.sh $MCR_DIRECTORY L 1 20 0.1 3 testfile.
Zip file Data 4: MATLAB® Toolkit for network traffic anomaly analysis. The toolkit performs: - De-trending and thresholding for anomaly detection; - Graph wavelet based summarizing and anomaly tracing; - Distribution fitting to spatial and temporal parameters; - Simulator/Emulator to regenerate statistically similar anomalies The toolkit is developed for Internet2 dataset, but customizable for other datasets.
Zip file Data 5: Source codes for Robust PCA experiments. MATLAB codes for Robust PCA related experiments. The codes generate random matrices and perform Robust PCA. The codes cover a range of decomposition and recovery experiments for RPCA.
Zip files 1-3 contain executable files; zip files 4-5 contain data and ReadMe files.
Department of Electrical and Computer Engineering
Rights Access
Subject
Internet traffic anomalies
Robust PCA
random matrices
anomaly modelling
Citation
Associated Publications
Bandara, Vidarshana W., Extraction, characterization and modeling of network data features - a compressive sensing and robust PCA based approach (Unpublished doctoral dissertation) Colorado State University, 2015. http://hdl.handle.net/10217/167001