Neural network architecture selection method and device and electronic equipment
A neural network and architecture technology, applied in neural architecture, neural learning methods, biological neural network models, etc., can solve problems such as time-consuming computing resources, complex target structure process, etc., and achieve the effect of rapid processing
Pending Publication Date: 2021-06-18
LENOVO (BEIJING) CO LTD
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0004] In view of this, the present application provides a neural network architecture selection method, which solves the problem that the process of finding the target structure in the NAS method in the prior art is complicated, consumes a lot of time and computing resources
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
[0052] The following will clearly and completely describe the technical solutions in the embodiments of the application with reference to the drawings in the embodiments of the application. Apparently, the described embodiments are only some of the embodiments of the application, not all of them. Based on the embodiments in this application, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the scope of protection of this application.
[0053] Neural network is a model in machine learning. It is an algorithmic mathematical model that imitates the behavior characteristics of animal neural networks and performs distributed parallel information processing. This kind of network depends on the complexity of the system, and achieves the purpose of processing information by adjusting the interconnection relationship between a large number of internal nodes.
[0054] The neural network architecture refers to the structure ...
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 neural network architecture selection method, which comprises the following steps: establishing a hidden feature space based on an architecture in an architecture pool, performing sampling in a high-precision first region in the hidden feature space, decoding hidden features in a sampling result to obtain a new architecture, and selecting a target structure from the new architecture. According to the method and device, the hidden features of the first interval with high precision in the hidden feature space are sampled, and the new architecture with higher precision is obtained based on hidden feature decoding, so that the target structure is selected based on precision, and compared with the prior art that the target structure can only be searched in a random selection mode in massive architectures in candidate architectures, the method and device have the advantage that the treatment process is quicker.
Description
technical field [0001] The present application relates to the field of artificial intelligence, and more specifically, relates to a neural network architecture selection method, device and electronic equipment. Background technique [0002] The use of deep learning technology has achieved unprecedented success in tasks such as image classification and detection, speech recognition, and natural language processing. In recent years, it has been widely used in many fields such as security, medical care, and advertising media. NAS (Neural Architecture Search, neural network architecture search) method began to show great potential in various fields. [0003] However, the design of the search space in the NAS method is more complicated. The NAS algorithm needs to sample randomly from millions or tens of millions of candidate architectures to find the target structure to achieve high performance, which consumes a lot of time and computing resources. Contents of the invention ...
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): G06N3/04G06N3/08
CPCG06N3/08G06N3/045
Inventor 郑欣悦王鹏尚雨薇
Owner LENOVO (BEIJING) CO LTD