Method of Analysis of Visualised Data

• US 20180165845A1
• Filed: 03/28/2017
• Published: 06/14/2018
• Est. Priority Date: 12/09/2016
• Status: Abandoned Application

• Current Assignee: Free Construction Sp. Z o.o.
• Original Assignee: Free Construction Sp. Z o.o.
• Inventors: Piolun-Noyszewski, Marcin, Weiss, Marcin, Sapala, Kamil
• Application Number:
  US15/471,951
• Publication Number:
  US 20180165845A1
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/00 Pattern recognition
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06T 11/206 Drawing of charts or graphs

Abstract

The object of the invention is a method of analysis of visualised data, comprising the steps of: data reception, statistical processing of data, generation of a pattern representing the processed results of analyses, further comprising detection of pattern features by means of a neural network, and grouping of result visualisations.

Claims

1. A method of analysis of visualised data, comprising the steps of:

data reception,

statistical processing of data,

generation of a pattern representing the processed results of analyses,

characterised in that it further comprises;

detection of pattern features by means of a neural network, and

grouping of result visualisations.

2. The method according to claim 1, characterised in that generation of a pattern consists in visualisation in the form of graphs.

3. The method according to claim 2, characterised in that the graphs belong to the groups of:

linear graphs, bar graphs, scatter-plots, box-plots, heat diagrams.

4. The method according to claim 1, characterised in that detection of pattern features consists in the analysis of graphs.

5. The method according to claim 2, characterised in that the graphs are converted to a pattern form.

6. The method according to claim 1, characterised in that the neural network is a convolutional network.

7. The method according to claim 6, characterised in that during detection of pattern features, the learned schemes are transferred between similar problems and the learned neural network is used for patterns of various resolutions.

8. The method according to claim 1, characterised in that the neural network is based on a graphics engine.

9. The method according to claim 1, characterised in that classification of a result to a particular group determines sending a signal to an external system.

10. The method according to claim 9, characterised in that the external system, as a result of the received signal, allows to block the user accessing to a particular informatic system.

11. The method according to claim 9, characterised in that the external system, as a result of the received signal, allows to block the user opening of a particular lock.

12. The method according to claim 9, characterised in that the external system, as a result of the received signal, allows execution of a particular security policy (set of rules) in the external system.

13. The method according to claim 3, characterised in that the graphs are converted to a pattern form.

14. The method according to claim 4, characterised in that the graphs are converted to a pattern form.