US 11,900,251 B2 · Amplification of initial training data

Abstract

Techniques are disclosed relating to increasing the amount of training data available to machine learning algorithms. A computer system may access an initial set of training data that specifies a plurality of sequences, each of which may define a set of data values. The computer system may amplify the initial set of training data to create a revised set of training data. The amplifying may include identifying sub-sequences of data values in ones of the plurality of sequences in the initial set of training data and using an inheritance algorithm to create a set of additional sequences of data values, where each one of the set of additional sequences may include sub-sequences of data values from at least two different sequences in the initial set of training data. The computer system may process the set of additional sequences using the machine learning algorithm to train a machine learning model.

Claims20

Dependent claims

A method, comprising:

accessing, by a computer system, an initial set of training data usable by a machine learning algorithm to train a machine learning model, wherein the initial set of training data comprises a set of sequences, wherein each sequence of the set of sequences defines a plurality of values in an order identifying a particular trend;
generating, by the computer system, one or more additional sequences to supplement the initial set of training data, wherein generating the one or more additional sequences includes generating at least one additional sequence of the one or more additional sequences by:
concatenating values of a first sequence of the set of sequences specified in the initial set of training data that are within a first sub-sequence with values of a second sequence of the set of sequences that are within a second sub-sequence;
and
processing, by the computer system, the initial set of training data and the one or more additional sequences using the machine learning algorithm to train the machine learning model.

2 more elements

A non-transitory computer-readable medium having program instructions stored thereon that are executable by a computer system to perform operations comprising:

accessing an initial set of training data usable by a machine learning algorithm to train a machine learning model, wherein the initial set of training data comprises a set of sequences, wherein each sequence of the set of sequences defines a plurality of values in an order identifying a particular trend;
generating one or more additional sequences to supplement the initial set of training data, wherein generating the one or more additional sequences includes generating at least one additional sequence of the one or more additional sequences by:
concatenating values of a first sequence of the set of sequences specified in the initial set of training data that are within a first sub-sequence with values of a second sequence of the set of sequences that are within a second sub-sequence;
and
processing the initial set of training data and the one or more additional sequences using the machine learning algorithm to train the machine learning model.

2 more elements

Dependencies are inferred from the claim text.

Amplification of initial training data

Life of this patent

Amplification of initial training data

Life of this patent

Abstract

Claims20