Looking for breakthrough ideas for innovation challenges? Try Patsnap Eureka!

Quantum average pooling calculation method

A computing method and quantum technology, applied in the field of quantum neural network computing, can solve problems such as incomparable supercomputing capabilities, and achieve the effects of reducing parameter dimensions, efficient storage and parallel processing

Pending Publication Date: 2020-06-26
HUBEI NORMAL UNIV
View PDF0 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0005] The purpose of the present invention is to propose a quantum average pooling calculation method for the above situation. This calculation method has supercomputing capabilities unmatched by classical computers, and provides solutions for large-scale calculation problems such as artificial intelligence and deep learning convolutional neural networks.

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
  • Quantum average pooling calculation method
  • Quantum average pooling calculation method
  • Quantum average pooling calculation method

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0045] Example 1: 4-qbit quantum information average pooling calculation:

[0046] Such as Figure 9As shown, the length and width of the convolution result information are 2-qbit each, and contain a total of 4-qbit quantum information, which is expanded by rows and recorded as the quantum state|f 4 >,|f 4 >=(c 0 ,c 1 ,c 2 ,c 3 ,c 4 ,c 5 ,c 6 ,c 7 ,c 8 ,c 9 ,c 10 ,c 11 ,c 12 ,c 13 ,c 14 ,c 15 ) T . Use the set average pooling kernel to perform an average pooling operation on the convolution result information, and the output contains a total of 2-qbit average pooling results, denoted as |g 2 >=(d 0 , d 1 , d 2 , d 3 ) T , take some values ​​for illustration, namely

[0047] In order to obtain the above average pooling result|g 2 >, the quantum state |f 4 > enter as Figure 7 In the quantum average pooling circuit model shown, the quantum state |f 4 > first via exchange gate D 4 After the action, the position of the ground state of the quantum st...

Embodiment 2

[0051] Example 2: 6-qbit quantum information average pooling calculation:

[0052] The convolution result information of each 3-qbit length and width|f 6 > After the average pooling kernel operation, the pooling result information is obtained|g 4 >, the schematic diagram of quantum average pooling is as follows Figure 12 , take part of the value to illustrate that is

[0053] First, the quantum state |f 6 >=(c 0 ,c 1 ,c 2 ,c 3 ,c 4 ,c 5 ,c 6 ,c 7 ,...,c 60 ,c 61 ,c 62 ,c 63 ) T Input into Quantum Circuit Model Figure 10 In , the position of the ground state of the quantum state is changed through the operation of the switch gate of 3 qubits, and the quantum state after the change is (c 0 ,c 1 ,c 8 ,c 9 ,c 2 ,c 3 ,c 10 ,c 11 ,...,c 54 ,c 55 ,c 62 ,c 63 ) T , the specific truth table of the exchange gate operation is as follows Figure 11 as shown in the table. At the next moment, the exchanged quantum state passes through the H gate and the id...

Embodiment 3

[0054] Embodiment 3: 8-qbit quantum information average pooling calculation: the calculation method is the same as the calculation method of embodiment 1 and embodiment 2, and the 8-qbit quantum average pooling circuit model is as follows Figure 13 shown.

[0055] Through the calculation process of the above-mentioned embodiment 1, embodiment 2 and embodiment 3, it can be extended to the 2n-qbit quantum average pooling circuit model, such as image 3 As shown, the length and width of the quantum convolution result information are each n-qbit, and expanded by row to get |f 2n >=(c 0 ,c 1 ,c 2 ,...,c 2n ) T As input information, the following steps are required to obtain the quantum average pooling result:

[0056] 1), the quantum state|f 2n >Input into the built quantum average pooling circuit model image 3 middle.

[0057] 2), through the exchange gate D 2n role, |f 2n > The position of the quantum state ground state changes.

[0058] 3) The first qubit and the 2...

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

No PUM Login to View More

Abstract

The invention relates to a quantum average pooling calculation method. The method comprises the following steps: expanding quantum convolution result information | fn > length and width n-qbit by rowto obtain | f2n > = (c0, c1, c2,..., c2n) T as input information, and inputting a quantum state | f2n > into a built quantum average pooling line model; under the action of an exchange gate D2n, keening the position of the absolute value of f2n > the quantum state ground state unchanged; performing H gate action on the first quantum bit and the 2nth quantum bit, performing unit matrix action on the rest quantum bits, and performing operating at the moment to obtain a quantum state ground state of addition, subtraction and addition, which represents a tensor product; outputting a quantum average pooling result through the acted quantum state measurement when the measurement result is an | 00 > state. The calculation method provided by the invention has a super-calculation capability which cannot be compared by a classic computer, and provides a solution for large-scale calculation problems such as artificial intelligence and deep learning convolutional neural networks.

Description

technical field [0001] The invention relates to the technical field of quantum neural network computing in the fields of speech analysis and image recognition, in particular to a quantum average pooling computing method. Background technique [0002] Convolutional neural network is a special kind of artificial neural network, which has become one of the most commonly used tools in the field of speech analysis and image recognition, so it is also a research hotspot of artificial neural network. Hubel D H et al. found that each single neuron does not respond to the entire image by studying the cat's visual cortex, but only responds to its responsible receptive field. Fukushima K and others first proposed the Neocognitron model based on the concept of receptive field, which is recognized as the first convolutional neural network (CNN) model. Lecun Y et al. proposed the LeNet-5 model, which consists of a convolutional layer, a downsampling layer, and a fully connected layer to ...

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
no application Login to View More
IPC IPC(8): G06N3/04G06N3/08G06N10/00
CPCG06N3/0463G06N3/082G06N10/00G06N3/045
Inventor 刘兴云闫茜茜王鹏程
Owner HUBEI NORMAL UNIV
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Patsnap Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Patsnap Eureka Blog
Learn More
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com