Unlock instant, AI-driven research and patent intelligence for your innovation.
System and method for identifying associations and evolution patterns among data elements
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A data and pattern technology, applied in the field of data analysis, can solve problems such as inability to obtain time series correlations, manual division, etc., and achieve the effects of optimizing time length and time lag constraints, improving accuracy, and identifying accurately
Active Publication Date: 2020-06-02
NEC CORP
View PDF2 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0006] However, the above method has the following problems: it can only mine the static association relationship in the time series, and it is necessary to manually divide the time series to obtain the association of different time periods
In addition, time-varying associations in time series cannot be obtained
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0034] The preprocessing unit 110 preprocesses the collected data to obtain a processed data sequence. In this embodiment, the collected data may include, for example, meteorological data and air quality data, and preferably also includes at least one item of traffic data, population density data, and pollutionsource data.
[0035] In this embodiment, the preprocessing performed by the preprocessing unit 110 may include the following aspects:
[0036] (1) Normalize the collected data into data sequences with the same time scale.
[0037] (2) Stabilize the normalized data sequence. For example, if the time series is non-stationary, it can be differenced to make it stationary.
[0038] (3) Based on the duration and time lag constraints, the smoothed data sequence is converted into a sample sequence. For example, the transformation parameters of each data element may be duration and time lag constraints. Transformation parameters may be set based on experience or prior knowl...
Embodiment 2
[0062] The difference between Embodiment 2 and Embodiment 1 is that: the duration and time lag constraints are optimized to improve the accuracy of causality discovery. This operation can be performed by figure 1 The preprocessing unit 110 shown in is implemented. Below, combine figure 1 The specific details of Embodiment 2 are described in detail with FIGS. 8-9.
[0063] First, set L>T according to empirical knowledge, that is, select an optimal time segment [t1, t2] in the range [1, L]. For example, L=10 may be set.
[0064] Then, use the following formula to get β:
[0065]
[0066] The formula includes two function penalty items, the former is a sparse penalty item, and the latter is a penalty item to ensure continuity. The meanings of the parameters in this formula are as follows:
[0067] -N is the number of samples, which can be regarded as the length of the Y sequence.
[0068] -L is a preset value, that is, in the [1, L] time range, select an optimal time se...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
PUM
Login to View More
Abstract
The invention provides a system for recognizing correlations among data elements of multivariate time sequence data and an evolution pattern. The system includes: a pre-treatment unit configured to perform pre-treatment on acquired data to obtain a treated data sequence; a correlation discovery unit configured to discover the correlations among the data elements of the multivariate time sequence data from the pre-treated data through a Granger causal graphical model to obtain a causal graphical sequence; and an evolution pattern discovery unit configured to perform aggregation on the obtained causal graphical sequence to discover an evolution pattern of a causality. The invention also provides a method for recognizing the correlations among the data elements of the multivariate time sequence data and the evolution pattern. Through the system and the method, the correlations among the data elements of the multivariate time sequence data and the evolution pattern can be effectively recognized, and the recognition of the causality is more accurate.
Description
technical field [0001] The present application relates to the field of data analysis, in particular to a system and method for identifying associations and evolution patterns among data elements of multivariate time series data. Background technique [0002] Identifying associations and evolution patterns among data elements of multivariate time-series data is a complex process because individual time-series data data elements change dynamically over time and space. Existing identification methods usually divide the time series into fixed-length segments based on experience, or manually set parameters based on prior knowledge. [0003] For example, Non-Patent Document 1 (“Learning Spatial-Temporal Varying Graphs with Applications to Climate Data Analysis”, Xi Chen, AAAI, 2010) proposes a pattern recognition method for dynamic relationship graphs based on Markov Random Field. The method mainly includes the following steps: [0004] Data preprocessing was performed first, wh...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More
Application Information
Patent Timeline
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to View More 
Login to View More