High-dimensional data visualization method supporting topological structure maintenance
A high-dimensional data and topology technology, applied in other database browsing/visualization, neural learning methods, other database retrieval, etc., can solve problems such as the influence of visualization results and the density distribution of difficult data sets
Active Publication Date: 2020-04-03
NANJING UNIV
View PDF5 Cites 3 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, the low-dimensional grid structure in the self-organizing neural network must be designed in advance, and the unreasonable grid structure has a great impact on the visualization results
No matter how many high-dimensional data points are mapped to the grid, each grid in the self-organizing neural network will only be labeled once, which makes it difficult to observe the density distribution of the data set in the low-dimensional grid
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 moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment
[0107] The embodiment of the present invention discloses a high-dimensional data visualization method that supports topology preservation, named as a high-dimensional data visualization method based on self-organizing incremental learning neural network, which is suitable for visualizing high-dimensional data, and in the visualization process To achieve topology preservation. The invention adaptively sets the network structure in the visible space, including the network shape and the number of reference points, and displays the high-dimensional data density distribution in the visible space.
[0108] Flow process of the present invention sees figure 1 . The visualization process of the present invention mainly includes several parts such as online clustering, visual mapping, and visual rendering. In the present invention, the main problem we consider is to map the data set X in the high-dimensional space to the visible space (two-dimensional space). Taking the 784-dimension...
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 high-dimensional data visualization method supporting topological structure maintenance, and particularly relates to a high-dimensional data visualization method based on a self-organizing incremental learning neural network, so as to realize self-adaptive network structure setting in a topological structure maintenance process and realize display of a visual spatial datadensity distribution condition. The method mainly comprises the steps of online clustering, visual mapping, visual rendering and the like. Online clustering is performed to learn a representative dataset capable of representing the structure from an original data set in a self-organizing manner; and according to visual mapping, a dimension reduction method is used for multi-dimensional scaling, the data set is mapped into a visual space, and the relative distance representing data is kept; and the distribution condition of the data set in the visual space is visually rendered and displayed. By adopting the method to perform high-dimensional data visualization, self-adaptive visual network structure generation can be realized, and the data density distribution condition is displayed whilethe data correlation is displayed in a visual result.
Description
technical field [0001] The invention relates to the field of high-dimensional data visualization, in particular to a high-dimensional data visualization method supporting topology structure maintenance. Background technique [0002] Today, advances in computer hardware technology make storing data fast and easy. Data in almost every field of daily life will be recorded, such as shopping website consumption records, mobile phone communication records, WeChat chat records, and so on. These recorded information constitute high-dimensional data. These high-dimensional data are like a storage pool of available information. Discovering potential information in high-dimensional data and extracting valuable information can help people better grasp laws and analyze trends. However, extracting valuable hidden information from high-dimensional data is not a trivial task. Simply using a computer to analyze high-dimensional data is difficult to be comprehensive. In order to better ex...
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
Login to View More Patent Type & Authority Applications(China)
IPC IPC(8): G06F16/904G06F16/901G06F16/906G06N3/04G06N3/08
CPCG06F16/904G06F16/9024G06F16/906G06N3/04G06N3/08
Inventor 窦慧申富饶徐百乐
Owner NANJING UNIV