Skip to content

fazzz/anom_ssl2

Repository files navigation

Anomaly detection

This is a tool based on Sparse Structure Learning (Ide et. al. 2008) for Givin Two Data sets of Time Serie

Requirement

LAPACK (https://github.com/Reference-LAPACK/lapack/archive/refs/tags/v3.10.0.tar.gz)

Installation

./configure LDFLAGS=-$LAPACK/lib make && make install

How to use

assl2 -Sparsity *r* -ndim *ndist* *time_series_A* *time_series_B* > *out*

Options

The followings are the options for assl2.

[-mode ] if 0 normal standarization will be done, if 1 not

[-ninia] initial step of data A

[-ninib] initial step of data B

[-nfina] final step of data A

[-nfinb] final step of data B

[-Sparsity] sparsity (from 0 to 1)

[-ndim] dimensionality of data

Output

Program Name:assl2
Data A is <time series A text file>
Data B is <time series B text file>

Number of the data set A = <N>
Number of the data set B = <N>
Options
Mode                        =  <0 or 1>
Initial step of data A      =  <n1>
Initial step of data B      =  <n2>
Final   step of data A      =  <N1>
Final   step of data B      =  <N2>
egrees of freedom           =  <M>
alue of sparsity parameter  =  <r>          

The block coordinate descent method will be started.
  2-th max diff:  0.020183
  3-th max diff:  0.000622
  4-th max diff:  0.000005
  5-th max diff:  0.000000
  6-th max diff:  0.000000
  7-th max diff:  0.000000
The block coordinate descent method is converged til   8-th iteration.
The block coordinate descent method will be started.
  2-th max diff:  0.017372
  3-th max diff:  0.000449
  4-th max diff:  0.000006
  5-th max diff:  0.000000
  6-th max diff:  0.000000
The block coordinate descent method is converged til   7-th iteration.
The Sparse structure of data set A
  1   2
  1   3
  1   4
  1   5
  1   6
  1   9
  2   3
  2   4
  2   5
  2   6
  2   9
...
The Sparse structure of data set B
  1   2
  1   3
  1   4
  1   5
  3   4
  3   5
  3  13
  4   5
  4   6
  4   7
...
The anomaly of each dimension
  #      a->b     b->a     max)
  1    0.000    0.000    0.000
  2    0.000    0.000    0.000
  3    0.000    0.000    0.000
  4    0.000    0.000    0.000
  5    0.001    0.001    0.001
  6    0.001    0.001    0.001
  7    0.000    0.000    0.000
  8    0.000    0.000    0.000
  9    0.000    0.000    0.000
 10    0.000    0.000    0.000

This output consists of several blocks, each presenting different information:

The first block displays the input file.

The second block shows the contents of the options used.

The third block presents the convergence of the computation.

The block "The sparse structure of data set A" illustrates the sparse correlation of data set A. The paired numbers represent the indices of degrees of freedom that exhibit correlation. Similarly, the block "The sparse structure of data set B" conveys the same meaning for data set B.

Finally, the block "The anomaly of each dimension" identifies the degrees of freedom that exhibit anomalies by comparing data set A and data set B. The first column represents the index of the degree of freedom, and when the value in the fourth column is nonzero, that degree of freedom is considered to have an anomaly.

About

"Application of Anomaly Detection to Identify Important Features of ProteinDynamics", Yu Yamamori* and Kentaro Tomii, ACS Omega(2025), CiteThis:https://doi.org/10.1021/acsomega.4c11546

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

 
 
 

Contributors