Jun 03

Nu-Anomica

by NDC-efoughty

Nu-Anomica is a One Class Support Vector Machine algorithm (OCSVM) that can be utilized on continuous (eg, engine temperature or fuel flow) data. A OCSVM takes a set of data, and for each new input in the data predicts which category the input might lie in (for instance: Anomaly, Not an Anomaly). However, many of the data sets involved in aviation are huge, and classical OCSVM’s can take a long time to run. Nu-Anomica has a much faster execution time compared to the classical version of OCSVM and yet produces a very similar outcome.

To utilize nu-Anomica, users will need to set up the data in the following way: Load your multivariate (train, validation and test) data in the workspace, where rows represent instances/observations and columns represent parameters/features. You can set the model parameters in the variable named “param”. In the nu-Anomica package there is a demo file (demoFile.m) that shows these steps. Follow the steps defined in the demoFile.m and the instructions in readme.txt to run the code. To execute, run the runAlgorithm.m file. For more details, please see the following paper: "nu-Anomica: A Fast Support Vector based Novelty Detection Technique," Santanu Das, Kanishka Bhaduri, Nikunj C. Oza and Ashok N. Srivastava

Nu-Anomica can be found here

Posted Jun 03, 2011 by NDC-efoughty

comments

Discussions

Add New Comment

  • guoyi-li 6 years, 10 months

    Dear Dr. Das,

    My name is Guoyi Li and I am currently a PHD student at Arizona State University working on a NASA project. I am studying your nu-Anomica code in Macbook system from nasa-Dashlink. When I run the demo, I got an error as following,

    Undefined function or variable 'mexSVM'.

    Error in SVMTrain (line 123)
    [AlphaY, SVs, Bias, Parameters, nSV, nLabel, mixM] = mexSVM(Samples, Labels, Parameters);

    Error in one_RbfSVC (line 50)
    [AlphaY, SVs, Bias, Parameters, nSV, nLabel] = SVMTrain(Samples, Labels, Parameters);

    Error in mainsvm (line 27)
    [AlphaY, SVs, Bias, Parameters, nSV, nLabel] =one_RbfSVC(Samples, Gamma,nu);

    Error in runAlgorithm (line 143)
    [SVs,score,nR,trtime,tsttime]= mainsvm(data,Sigma_in,nu);

    Error in demoFile (line 29)
    [ocsvm] =runAlgorithm(trdata,vlddata,xx,'ocsvms','saveOneClass',params);

    I noticed that you modified the 'SVMTrain.m' in OSU-SVM source code as following,
    [AlphaY, SVs, Bias, Parameters, nSV, nLabel, mixM] = mexSVM(Samples, Labels);

    and made an additional package RoutineCall/MexFile. I am wondering if you could kindly give me any suggestions to address this issue. I appreciate your time and thank you very much in advance!

    Sincerely,
    Guoyi

  • Reply to comment

Subscribe

Archive

2011