Training data management method and device based on block chain
A technology for training data and management methods, applied in the computer field, can solve problems such as inability to perform specific tasks well, poor training data, adding intelligent robots, etc., and achieve the effect of avoiding malicious changes, high quality, and avoiding self-learning
Active Publication Date: 2019-12-03
LI ZHAN SHANGHAI TECH CO LTD
View PDF14 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0003] If the quality of the training data itself is poor, or the training data is maliciously changed by an intruder and the quality is poor, after the intelligent robot learns independently based on the poor-quality training data, it may not be able to complete specific tasks well, and even increase the occurrence of other problems for the intelligent robot risks of
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 Construction
[0041] The training data used for autonomous learning of intelligent robots usually needs to be able to indicate the specific environment that the intelligent robot may exist when performing a specific task, and the possible response mode that may exist when the intelligent robot performs a specific task in a specific environment.
[0042] For the low-quality training data generated due to excessive noise in the data collection process, it may cause the intelligent robot after autonomous learning based on the training data to be unable to perform specific tasks well. For example, it is expected that the intelligent robot can have the ability to play badminton with a real human object. If the training data indicates that the intelligent robot performs the task of playing badminton with a real human object, the specific environment that may exist includes "the shuttlecock is played at a set time. The section moves with a certain trajectory", and indicates that in this specific en...
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 embodiment of the invention provides a training data management method and device based on a block chain. The method comprises the steps of obtaining candidate training data; regulating and controlling the intelligent robot to respond in a response mode indicated by the candidate training data in a specific environment indicated by the candidate training data, obtaining a corresponding response result, and the response result indicating the quality of the candidate training data; under the condition that a response result meets a preset quality requirement, storing the candidate training data as training data for autonomous learning of the intelligent robot to a block chain; and in response to the triggering condition, triggering the intelligent robot to autonomously learn according tothe training data stored in the block chain. According to the technical scheme of the embodiment of the invention, autonomous learning of the intelligent robot according to training data with poor quality can be avoided more effectively.
Description
technical field [0001] This description relates to the field of computer technology, in particular to a blockchain-based training data management method and device. Background technique [0002] With the rapid development of artificial intelligence technology, intelligent robots with autonomous learning capabilities continue to emerge. Intelligent robots can learn autonomously based on the training data provided to them, thereby improving their efficiency in completing specific tasks, or increasing their ability to complete specific tasks. [0003] If the quality of the training data itself is poor, or the quality of the training data is maliciously changed by an intruder, after the intelligent robot learns independently based on the poor-quality training data, it may not be able to complete specific tasks well, and even increase the occurrence of other problems for the intelligent robot. risks of. [0004] In view of this, it is hoped that there will be an improved solutio...
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): G06F16/27G06F16/35G06K9/62
CPCG06F16/27G06F16/35G06F18/214
Inventor 李立中李知旃
Owner LI ZHAN SHANGHAI TECH CO LTD