System and method for active machine learning
AssigneeSamsung Electronics Co., Ltd.
InventorsYue Deng, Yilin Shen, Hongxia Jin
Life of this patent
Estimating the term…
Abstract
An electronic device for active learning includes at least one memory and at least one processor coupled to the at least one memory. The at least one processor is configured to select one or more entries from a data set including unlabeled data based on a similarity between the one or more entries and labeled data. The at least one processor is further configured to cause the one or more entries to be labeled.
Claims25
A method for active learning, the method comprising:
- selecting one or more entries from a data set including unlabeled data based on a similarity between the one or more entries and labeled data;
- and
- causing the one or more entries to be labeled.
An electronic device for active learning, the electronic device comprising:
- at least one memory;
- and
- at least one processor coupled to the at least one memory, wherein the at least one processor is configured to:
- select one or more entries from a data set including unlabeled data based on a similarity between the one or more entries and labeled data;
- and
- cause the one or more entries to be labeled.
A non-transitory computer readable medium embodying a computer program for operating an electronic device including at least one processor, the computer program comprising computer readable instructions that, when executed by the at least one processor, cause the electronic device to:
- select one or more entries from a data set including unlabeled data based on a similarity between the one or more entries and labeled data;
- and
- cause the one or more entries to be labeled.
A method for training an active learning model, the method comprising:
- receiving a data pool including labeled data and unlabeled data;
- mapping, by a feature encoder, the labeled data and the unlabeled data to a latent space;
- and
- processing the mapped labeled data and the mapped unlabeled data by a discriminator, wherein the discriminator is configured to distinguish unlabeled data from labeled data.
Dependencies are inferred from the claim text.