Back to results
US 11,687,839 B2Granted
System and method for generating and optimizing artificial intelligence models
AssigneeActapio, Inc.
InventorShin-ichiro Okamoto
Life of this patent
Estimating the term…
Abstract
A computer implemented method for generating and optimizing an artificial intelligence model, the method comprising receiving input data and labels, and performing data validation to generate a configuration file, and splitting the data to generate split data for training and evaluation; performing training and evaluation of the split data to determine an error level, and based on the error level, performing an action, wherein the action comprises at least one of modifying the configuration file and tuning the artificial intelligence model automatically; generating the artificial intelligence model based on the training, the evaluation and the tuning; and serving the model for production.
Claims20
A computer implemented method for generating and optimizing an artificial intelligence model, the method comprising:
- receiving input data and labels, and performing data validation to generate a configuration file, and splitting the data to generate split data for training and evaluation, wherein the configuration file comprises a schema file, a feature file, a model file, and a cross feature file for generating the artificial intelligence model;
- performing training and evaluation of the split data to determine an error level, and based on the error level, performing an action, wherein the action comprises at least one of modifying the configuration file and tuning the artificial intelligence model automatically;
- generating the artificial intelligence model based on the training, the evaluation and the tuning;
- and
- serving the model for production,
- wherein the performing data validation comprises:
- for the input data being integer type data:
- determine a data density based on a ratio of unique count with respect to maximum value count of data,
- for the data density being greater than or equal to a first threshold:
- characterizing the input data as contiguous;
- determining if the maximum value count of data is greater than or equal to a second threshold value;
- for the maximum value count of data being greater than or equal to the second threshold value, characterizing the input data as contiguous and large, and providing the input data with categorical column with identity and embedding column;
- and
- for the maximum value count of data being smaller than the second threshold value, characterizing the input data as contiguous and small, and providing the input data with categorical column with identity;
- for the density being less than the first threshold:
- characterizing the input data as sparse;
- comparing the unique count against a third threshold;
- for the unique count being greater than or equal to the third threshold, characterizing the input data as large and sparse, and providing the input data with categorical column with hash bucket and embedding column;
- and
- for the unique count being less than the third threshold, characterizing the input data as small and sparse, and providing the input data with categorical column with hash bucket, and
- generating a validation report, wherein the validation report provides information on data errors associated with the input data.
18 more elements
A non-transitory computer readable medium configured to execute machine-readable instructions stored in a storage, for generating and optimizing an artificial intelligence model, the instructions comprising:
- receiving input data and labels, and performing data validation to generate a configuration file, and splitting the data to generate split data for training and evaluation, wherein the configuration file comprises a schema file, a feature file, a model file, and a cross feature file for generating the artificial intelligence model;
- performing training and evaluation of the split data to determine an error level, and based on the error level, performing an action, wherein the action comprises at least one of modifying the configuration file and tuning the artificial intelligence model automatically;
- generating the artificial intelligence model based on the training, the evaluation and the tuning;
- and
- serving the model for production,
- wherein the performing data validation comprises:
- for the input data being integer type data:
- determine a data density based on a ratio of unique count with respect to maximum value count of data,
- for the data density being greater than or equal to a first threshold:
- characterizing the input data as contiguous;
- determining if the maximum value count of data is greater than or equal to a second threshold value;
- for the maximum value count of data being greater than or equal to the second threshold value, characterizing the input data as contiguous and large, and providing the input data with categorical column with identity and embedding column;
- and
- for the maximum value count of data being smaller than the second threshold value, characterizing the input data as contiguous and small, and providing the input data with categorical column with identity;
- for the density being less than the first threshold:
- characterizing the input data as sparse;
- comparing the unique count against a third threshold;
- for the unique count being greater than or equal to the third threshold, characterizing the input data as large and sparse, and providing the input data with categorical column with hash bucket and embedding column;
- and
- for the unique count being less than the third threshold, characterizing the input data as small and sparse, and providing the input data with categorical column with hash bucket, and
- generating a validation report, wherein the validation report provides information on data errors associated with the input data.
18 more elements
A system for generating and optimizing an artificial intelligence model, the system comprising:
- a data framework configured to receive input data and labels, perform data validation to generate a configuration file, split the data to generate split data for training and evaluation;
- a deep framework configured to perform training and evaluation of the split data to determine an error level, and based on the error level, to perform an action, generate the artificial intelligence model based on the training, the evaluation and the tuning, and serve the model for production;
- and
- a tuning framework configured to perform the action, wherein the action comprises at least one of modifying the configuration file and tuning the artificial intelligence model automatically,
- wherein the configuration file comprises a schema file, a feature file, a model file, and a cross feature file for generating the artificial intelligence model in the deep framework, and
- wherein the performing data validation comprises:
- for the input data being integer type data:
- determine a data density based on a ratio of unique count with respect to maximum value count of data,
- for the data density being greater than or equal to a first threshold:
- characterizing the input data as contiguous;
- determining if the maximum value count of data is greater than or equal to a second threshold value;
- for the maximum value count of data being greater than or equal to the second threshold value, characterizing the input data as contiguous and large, and providing the input data with categorical column with identity and embedding column;
- and
- for the maximum value count of data being smaller than the second threshold value, characterizing the input data as contiguous and small, and providing the input data with categorical column with identity;
- for the density being less than the first threshold:
- characterizing the input data as sparse;
- comparing the unique count against a third threshold;
- for the unique count being greater than or equal to the third threshold, characterizing the input data as large and sparse, and providing the input data with categorical column with hash bucket and embedding column;
- and
- for the unique count being less than the third threshold, characterizing the input data as small and sparse, and providing the input data with categorical column with hash bucket, and
- generating a validation report, wherein the validation report provides information on data errors associated with the input data.
18 more elements
Dependencies are inferred from the claim text.