System and method for detecting outliers in real-time for a univariate time-series signal

Patent metadata

Field	Value
Publication number	US-10733264-B2
Application number	US-201615184494-A
Country	US
Kind code	B2
Filing date	Jun 16, 2016
Priority date	Jun 17, 2015
Publication date	Aug 4, 2020
Grant date	Aug 4, 2020

Abstract

Disclosed is a method and system for detecting outliers in real-time for a univariate time-series signal. The system may receive the univariate time-series signal, comprising a plurality of datasets, from a data source. The system may compute a standard deviation of a dataset of the plurality of datasets. Subsequently, the system may compute the optimal sample block size and the critical sample size of the dataset. Further, the system may determine the optimal operational block size of the dataset. The system may segment the plurality of datasets into blocks based upon the optimal operational block size. The system may detect the outliers by performing an outlier detection technique on the blocks, thereby ensuring improved execution time while minimally affecting precision and accuracy of the outcome of the outlier detection method.

First claim

What is claimed is:

1. A method for detecting outliers in real-time for a univariate time-series signal, the method comprising:
receiving, by a processor 210 , a univariate time-series signal from a data source having stored data captured by sensors, wherein the univariate time-series signal comprises a plurality of datasets, and wherein each dataset of the plurality of datasets comprises number of univariate time-series data elements;
computing, by the processor 210 , a standard deviation (σ) of a dataset of the plurality of datasets;
computing, by the processor 210 , an optimal sample block size ( ) of the dataset by using the standard deviation (σ);
computing, by the processor 210 , a critical sample size ( critical ) of the dataset based on the standard deviation (σ), the N number of univariate time-series data elements, an accuracy (δ) in minimizing false alarms, and a precision (1−∈) of outcome related to the critical sample size critical ;
determining, by the processor 210 , an optimal operational block size ( operational ) of the dataset using operational ={(| | mod critical =0}, wherein indicates number of univariate time-series data elements and critical indicates the critical sample size;
segmenting, by the processor 210 , the plurality of datasets into blocks based upon the optimal operational block size ( operational ), wherein each block comprises / operational data elements of the number of univariate time-series data elements; and
detecting, by the processor 210 , outliers in real-time by performing an outlier detection technique on the blocks of the segmented plurality of datasets, wherein the outlier detection technique comprises one or more unsupervised techniques including at least one of a Rosner filtering technique to minimize swamping effects or a Hampel filtering technique to minimize masking effects.

2. The method of claim 1 , wherein the optimal sample block size

=

σ

(

1
-
δ

)

⁢

ϵ
2

.

wherein σ indicates the standard deviation, ∈ indicates precision loss and δ indicates the accuracy.

3. The method of claim 1 , wherein the critical sample size ( critical ) is computed using

critical

=

min
⁢

{

σ

(

1
-
δ

)

⁢

ϵ
2

,


ℕ


/
2

}

.

wherein σ indicates the standard deviation, ∈ indicates precision loss, δ indicates the accuracy and indicates number of univariate time-series data elements.

4. A system implemented on a cloud-based environment for detecting outliers in real-time for a univariate time-series signal, the system comprises:
a processor 210 ;
a memory 212 coupled to the processor 210 , wherein the processor is capable for executing programmed instructions stored in the memory 212 to:
receive a univariate time-series signal from a data source having stored data captured by sensors, wherein the univariate time-series signal comprises a plurality of datasets, and wherein each dataset of the plurality of datasets comprises number of univariate time-series data elements;
compute a standard deviation (σ) of a dataset of the plurality of datasets;
compute an optimal sample block size ( ) of the dataset by using the standard deviation (σ);
compute a critical sample size ( critical ) of the dataset based on the standard deviation (σ), the number of univariate time-series data elements, an accuracy (δ) in minimizing false alarms, and a precision (1−∈) of outcome related to the critical sample size critical ;
determine an optimal operational block size ( operational ) of the dataset using operational ={(| | mod critical =0}, wherein indicates number of univariate time-series data elements and critical indicates the critical sample size;
segment the plurality of datasets into blocks based upon the optimal operational block size ( operational ), wherein each block comprises / operational data elements of the number of univariate time-series data elements; and
detect outliers in real-time by performing an outlier detection technique on the blocks of the segmented plurality of datasets, wherein the outlier detection technique comprises one or more unsupervised techniques including at least one of a Rosner filtering technique to minimize swamping effects or a Hampel filtering technique to minimize masking effects.

5. The system of claim 4 , wherein the sample block size ( ) is computed using

(
)

⁢

⁢
is
⁢

⁢
computed
⁢

⁢
using

=

σ

(

1
-
δ

)

⁢

ϵ
2

.

wherein σ indicates the standard deviation, ∈ indicates precision loss and δ indicates the accuracy.

6. The system of claim 4 , wherein the critical sample size ( critical ) is computed using

critical

=

min
⁢

{

σ

Assignees

• Tata Consultany Services Ltd
• Tata Consultancy Services Ltd

Inventors

• Ukil Arijit
• Bandyopadhyay Soma
• Pal Arpan

Classifications

• G06F17/18 Primary

  for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)} · CPC title

• H04W4/00

  Services specially adapted for wireless communication networks; Facilities therefor · CPC title