Shifter implemented circulant permutation matrix operations
US-2024386072-A1 · Nov 21, 2024 · US
US11687616B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-11687616-B2 |
| Application number | US-202017090940-A |
| Country | US |
| Kind code | B2 |
| Filing date | Nov 6, 2020 |
| Priority date | Dec 13, 2019 |
| Publication date | Jun 27, 2023 |
| Grant date | Jun 27, 2023 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
An arithmetic processing apparatus includes a memory and a processor. The processor coupled to memory and configured to determine an individual not to be evolved to an individual of a second generation from among a plurality of individuals in a first generation based on a predetermined reference for calculation completion of fitness calculation for each of the plurality of individuals, the second generation being a generation next to the first generation, and determine to cause the determined individual to evolve to an individual of a generation next or subsequent to the second generation.
Opening claim text (preview).
What is claimed is: 1. An arithmetic processing apparatus of calculating in an evolutionary computation that calculates fitnesses of a plurality of individuals by parallel processing, the plurality of individuals being based on a plurality of inputs, the apparatus comprising: a memory; and a processor circuit coupled to memory and configured to: use, in the processor circuit, a technique that, for calculating a fitness of an individual, performs prediction of an eigen solution by using a predictor represented by a neural network and starts a calculation for search of the eigen solution from a result of the prediction to acquire an accurate eigen solution; determine, in the processor circuit, an individual not to be evolved to an individual of a second generation from among a plurality of individuals in a first generation based on a predetermined reference for calculation completion of fitness calculation for each of the plurality of individuals, the second generation being a generation next to the first generation, the predetermined reference for calculation completion is one of a number of searches for searching an eigen solution of fitness calculation for the plurality of individuals in the first generation, a calculation time for searching an eigen solution which is taken for fitness calculation for a plurality of individuals in the first generation, a number of completed individuals of fitness calculation for a plurality of individuals in the first generation, or a proportion of eigen solution searches that are completed; and determine, in the processor circuit, to cause the determined individual to evolve to an individual of a generation next or subsequent to the second generation. 2. The arithmetic processing apparatus according to claim 1 , wherein the processor circuit is further configured to: determine, in the processor circuit, the individual not to be evolved to individuals of the second generation based on the number of searches for searching an eigen solution of fitness calculation for the plurality of individuals in the first generation, control, in the processor circuit, to continue the search for an eigen solution for the decided individuals, and determine, in the processor circuit, to evolve the individual to individuals of a generation that is a next or subsequent generation of the second generation. 3. The arithmetic processing apparatus according to claim 1 , wherein the processor circuit is further configured to: determine, in the processor circuit, the individual not to be evolved to individuals of the second generation based on the calculation time for searching an eigen solution, control, in the processor circuit, to continue the search for an eigen solution for the individual, and determine, in the processor circuit, to evolve the individual to individuals of a generation that is a next or subsequent generation of the second generation based on the continued search. 4. The arithmetic processing apparatus according to claim 1 , wherein the processor circuit is further configured to: determine, in the processor circuit, the individual not to be evolved to individuals of the second generation based on the number of completed individuals of fitness calculation for a plurality of individuals in the first generation, and control, in the processor circuit, to continue the search for an eigen solution for the individual, and determine, in the processor circuit, to evolve the decided individuals to individuals of a generation that is a next or subsequent generation of the second generation. 5. The arithmetic processing apparatus according to claim 1 , wherein the processor circuit is one of a plurality of processors, and when the number of a plurality of individuals in the first generation is higher than the number of the plurality of processors, each of the plurality of processors executes the fitness calculation for an individual for which the calculation has not been executed after completing the fitness calculation for one individual. 6. A non-transitory computer-readable storage medium storing a program that causes a processor circuit included in a computer to execute a processing of calculating in an evolutionary computation that calculates fitnesses of a plurality of individuals by parallel processing, the plurality of individuals being based on a plurality of inputs, the processing comprising: using, in the processor circuit, a technique that, for calculating a fitness of an individual, performs prediction of an eigen solution by using a predictor represented by a neural network and starts a calculation for search of the eigen solution from a result of the prediction to acquire an accurate eigen solution; determining, in the processor circuit, an individual not to be evolved to individual of a second generation from among a plurality of individuals in a first generation based on a predetermined reference for calculation completion of fitness calculation for each of the plurality of individuals, the second generation being a generation next to first generation, wherein the predetermined reference for calculation completion is one of a number of searches for searching an eigen solution of fitness calculation for the plurality of individuals in the first generation, a calculation time for searching an eigen solution which is taken for fitness calculation for a plurality of individuals in the first generation, a number of completed individuals of fitness calculation for a plurality of individuals in the first generation, or a proportion of eigen solution searches that are completed; and determining, in the processor circuit, to cause the determined individual to evolve to individuals of a generation next or subsequent to the second generation. 7. The non-transitory computer-readable storage medium according to claim 6 , wherein the processing further comprising: determining, in the processor circuit, the individual not to be evolved to individuals of the second generation based on the number of searches for searching an eigen solution of fitness calculation for the plurality of individuals in the first generation, controlling, in the processor circuit, to continue the search for an eigen solution for the decided individuals, and determining, in the processor circuit, to evolve the individual to individuals of a generation that is a next or subsequent generation of the second generation. 8. The non-transitory computer-readable storage medium according to claim 6 , wherein the processing further comprising: determining, in the processor circuit, the individual not to be evolved to individuals of the second generation based on the calculation time for searching an eigen solution, controlling, in the processor circuit, to continue the search for an eigen solution for the individual, and determining, in the processor circuit, to evolve the individual to individuals of a generation that is a next or subsequent generation of the second generation based on the continued search. 9. An arithmetic processing method of calculating in an evolutionary computation that calculates fitnesses of a plurality of individuals by parallel processing, the plurality of individuals being based on a plurality of inputs, the arithmetic processing method comprising: using, in a processor circuit, a technique that, for calculating a fitness of an individual, performs prediction of an eigen solution by using a predictor represented by a neural network and starts a calculation for search of the eigen solution from a result of the prediction to acquire an accurate eigen solution; determining, in the processor circuit, an individual not to be evolved to individual of a second generation from among a plurality of individuals
Feedforward networks · CPC title
Supervised learning · CPC title
Matrix or vector computation {, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization (matrix transposition G06F7/78)} · CPC title
Learning methods · CPC title
in which a parameter or coefficient is automatically adjusted to optimise the performance · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.