http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-107341522-A
Outgoing Links
Predicate | Object |
---|---|
assignee | http://rdf.ncbi.nlm.nih.gov/pubchem/patentassignee/MD5_ae2be58a477cef66248d0dc28cb0ea82 |
classificationCPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentcpc/G06F18-23213 |
classificationIPCInventive | http://rdf.ncbi.nlm.nih.gov/pubchem/patentipc/G06K9-62 |
filingDate | 2017-07-11-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
inventor | http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_c7bebd971cf97963cfe160abe44e9caa http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_66b18542833ccba25cd464f39db2bc28 http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_d55c1e1e7151d8dfb021eab440e5451f http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_6c0ec21f2802c417b9142ec156e5b6ee http://rdf.ncbi.nlm.nih.gov/pubchem/patentinventor/MD5_6d38037868ae913c1f8d85cb602eef89 |
publicationDate | 2017-11-10-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
publicationNumber | CN-107341522-A |
titleOfInvention | A method for unlabeled text and image recognition based on density semantic subspace |
abstract | The invention discloses a method for unlabeled text and image recognition based on a density semantic subspace, comprising the following sub-steps: S1 uses a Gaussian kernel function to estimate the sample density of the original space; S2 uses a Cauchy kernel function to estimate the sample density of the intrinsic semantic space Density; S3 minimizes the objective function, combines the sample density function of the original space and the density function of the intrinsic semantic space to obtain the objective function, uses the steepest descent method to minimize the objective function, and obtains the low-dimensional semantic space representation of the data; S4 Semantic space aggregation Classes, using the K-means algorithm to achieve clustering in the semantic space. The present invention makes the manifold structure of the data clear by learning the homeomorphic transformation from the high-dimensional manifold to the local volume preserving in the eigendimensional space, thereby effectively solving the small sample clustering problem. |
isCitedBy | http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-109783806-B http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-109783806-A http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-109446910-B http://rdf.ncbi.nlm.nih.gov/pubchem/patent/CN-109446910-A |
priorityDate | 2017-07-11-04:00^^<http://www.w3.org/2001/XMLSchema#date> |
type | http://data.epo.org/linked-data/def/patent/Publication |
Incoming Links
Total number of triples: 22.