| Predicate |
Object |
| assignee |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentassignee/MD5_d6a6f422b091ba12ea61d4adbf1b0e8e |
| classificationCPCAdditional |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F2119-12 |
| classificationCPCInventive |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-044
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F30-27
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-084
http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06N3-045 |
| classificationIPCAdditional |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06F119-12 |
| classificationIPCInventive |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06F30-27
http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06N3-08
http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06N3-04 |
| filingDate |
2020-12-30-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| inventor |
http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_a11edfc68ec7e25b6c10a7a016228187
http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_fabcbbe486bca0f2dc257a631fd8affb |
| publicationDate |
2021-04-30-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| publicationNumber |
CN-112733444-A |
| titleOfInvention |
Multi-step long time series prediction method based on CycleGAN neural network |
| abstract |
The invention relates to the field of time series prediction, and aims to provide a multi-step long time series prediction method based on CycleGAN neural network. Including: building a data set; building a deep neural network model based on CycleGAN, the model has a paired generative adversarial network structure, including two generators and two discriminators; where the generator is used to generate the real data to be predicted. Distribution, the discriminator is used to judge whether the generated data conforms to the real data distribution; train the network model, train the generator and the discriminator alternately in turn, and use the error to backpropagate to optimize the parameters; use the trained generator to predict and output forecast result. The invention uses the neural network technology to perform multi-step time series prediction, and can capture the high-dimensional statistical characteristics of the data through cyclic confrontation training to obtain high-precision prediction results. Applicable to more datasets than existing techniques. |
| isCitedBy |
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-114545255-A
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-113537586-A
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-113311703-A
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-115288994-A
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-115526300-B
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-113378655-A
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-115526300-A
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-113378655-B
http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-113180684-A |
| priorityDate |
2020-12-30-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
| type |
http://data.epo.org/linked-data/def/patent/Publication |