# US 2021/0279643 A1 · Method and system for generating best performing data models for datasets in a computing environment > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0279643 A1 - Document ID: US20210279643A1 - Status: Published application - Kind: A1 - Application number: 17/328,234 - Filed: 2021-05-24 - Published: 2021-09-09 - Earliest claimed: 2017-07-18 - Assignees: iQGateway LLC - Applicants: iQGateway LLC - Inventors: Praveen Koduru - Web page: https://patentgrep.com/patent/US20210279643A1 ## CPC - G06N 3/126: Evolutionary algorithms, e.g. genetic algorithms or genetic programming - G06F 11/328: {Computer systems status display (G06F11/327 takes precedence)} - G06F 11/3447: {Performance evaluation by modeling} - G06F 18/2148: {characterised by the process organisation or structure, e.g. boosting cascade} - G06F 18/285: {Selection of pattern recognition techniques, e.g. of classifiers in a multi-classifier system} - G06N 20/00: Machine learning - G06N 5/01: Dynamic search techniques; Heuristics; Dynamic trees; Branch-and-bound - G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract A method and system for generating best performing data models for datasets is disclosed. The method includes obtaining input dataset from external data sources. The method includes dividing dataset into three parts based on type of data variables and data samples. The method includes retrieving data modelling methods from a database and generating first set of data models for the first part of the dataset. The method includes determining first fitness function for the first part of the dataset based on the first set of data model. The method includes generating second set of data models for the second part of the dataset and determining second fitness function based on the second set of data model. Further, the method includes determining best performing data model among the first set of data models and the second set of data model. Additionally, the method includes outputting the best performing data model. ## Claims ### Claim 1 A method for generating best performing data models for datasets in a computing environment, wherein the method comprises: obtaining, by a processor, an input dataset from one or more external data sources, wherein the input dataset comprises data variables and data samples; dividing, by the processor, the obtained dataset into at least three parts based on a type of data variables and the data samples, wherein each part of the dataset comprises samples from the input dataset, and wherein a first part of the dataset comprises training data, a second part of the dataset comprises testing data, and a third part of the dataset comprises validation data; retrieving, by the processor, one or more data modeling methods from a pool of data modeling methods stored in a database, wherein the one or more data modeling methods comprises one or more parameter ranges; generating, by the processor, a first set of data models for the first part of the dataset by using the retrieved one or more data modeling methods, wherein an input to generate the first set of data models is the first part of the dataset; determining, by the processor, a first fitness function for the first part of the dataset based on the generated first set of data models; generating, by the processor, a second set of data models for the second part of the dataset by validating the first set of data models and the determined first fitness function with the second part of the dataset; determining, by the processor, a second fitness function associated with the second part of the dataset based on the generated second set of data models; determining, by the processor, at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function; and outputting, by the processor, the determined at least one best performing data model on a user interface of a user device. ### Claim 2 (depends on 1) The method of claim 1, wherein the one or more parameter ranges includes at least one of a set comprising prediction range, accuracy range, precision range, and recall range. ### Claim 3 (depends on 1) The method of claim 1, wherein the one or more data modelling methods comprises one or more machine learning methods and one or more artificial intelligence methods configured for generating data models. ### Claim 4 (depends on 1) The method of claim 1, wherein the first set of data models represent one or more data patterns inherent in the first part of the dataset and the second set of data models represents one or more data patterns inherent in the second part of datasets. ### Claim 5 (depends on 1) The method of claim 1, wherein determining the at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function comprises: determining the at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function using a pareto front method. ### Claim 6 (depends on 1) The method of claim 1, wherein determining the at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function comprises: determining performance of the first set of data models and the second set of data models based on the first fitness function and the second fitness function; and ranking the first set of data models and the second set of data models based on the determined performance of the first set of data models and the second set of data models. ### Claim 7 (depends on 1) The method of claim 1, further comprising: validating the determined at least one best performing data model using the third part of the dataset to generate a final best performing data model; and outputting the final best performing data model on the user interface of the user device. ### Claim 8 A system for generating best performing data models for datasets in a computing environment, the system comprises: one or more hardware processors; and a memory coupled to the one or more hardware processors, wherein the memory comprises a plurality of subsystems in the form of programmable instructions executable by the one or more hardware processors, wherein the plurality of subsystems comprises: a dataset receiver module configured for obtaining an input dataset from one or more external data sources, wherein the input dataset comprises data variables and data samples; a dataset segmentation module configured for dividing the obtained dataset into at least three parts based on a type of the data variables and the data samples, wherein each part of the dataset comprises samples from the input the dataset, and wherein a first part of the dataset comprises training data, a second part of the dataset comprises testing data, and a third part of the dataset comprises validation data; a data model method retrieval module configured for retrieving one or more data modeling methods from a pool of data modeling methods stored in a database, wherein the one or more data modeling methods comprises one or more parameter ranges; a data model generator module configured for generating a first set of data models for the first part of the dataset by using the retrieved one or more data modeling methods, wherein input to generate the first set of data models is the first part of the dataset; a fitness function determination module configured for determining a first fitness function for the first part of the dataset based on the generated first set of data models; the data model generator module further configured for generating a second set of data models for the second part of the dataset by validating the first set of data models and the determined first fitness function with the second part of the dataset; the fitness function determination module configured for determining a second fitness function associated with the second part of the dataset based on the generated second set of data models; a best performing data model determination module configured for determining at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function; and an output module configured for outputting the determined at least one best performing data model on a user interface of a user device. ### Claim 9 (depends on 8) The system of claim 8, wherein the one or more parameter ranges includes at least one of a set comprising prediction range, accuracy range, precision range, and recall range. ### Claim 10 (depends on 8) The system of claim 8, wherein the one or more data modelling methods comprises one or more machine learning methods and one or more artificial intelligence methods configured for generating data models. ### Claim 11 (depends on 8) The system of claim 8, wherein the first set of data models represent one or more data patterns inherent in the first part of the dataset and the second set of data models represents one or more data patterns inherent in the second part of the dataset. ### Claim 12 (depends on 8) The system of claim 8, wherein in determining the at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function, the best performing data model determination module is configured for: determining the at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function using a pareto front method. ### Claim 13 (depends on 8) The system of claim 8, wherein in determining the at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function, the best performing data model determination module is configured for: determining performance of the first set of data models and the second set of data models based on the first fitness function and the second fitness function; and ranking the first set of data models and the second set of data models based on determined performance of the first set of data models and the second set of data models. ### Claim 14 (depends on 8) The system of claim 8, wherein the best performing data model determination module is further configured for: validating the determined at least one best performing data model using the third part of the dataset to generate a final best performing data model; and outputting the final best performing data model on the user interface of the user device. ### Claim 15 A non-transitory computer-readable storage medium having instructions stored therein that, when executed by a hardware processor, causes the processor to perform method steps comprising: obtaining an input dataset from one or more external data sources, wherein the input dataset comprises data variables and data samples; dividing the obtained dataset into at least three parts based on a type of the data variables and the data samples, wherein each part of the dataset comprises an samples from the input datasets, and wherein the first part of the dataset comprises training data, the second part of the dataset comprises testing data, and third part of the dataset comprises validation data; retrieving one or more data modelling methods from a pool of data modelling methods stored in a database, wherein the one or more data modelling methods comprises one or more parameter ranges; generating a first set of data models for the first part of the dataset by using the retrieved one or more data modelling methods, wherein input to generate the first set of data models is the first part of the dataset; determining a first fitness function for the first part of the dataset based on the generated first set of data models; generating the second set of data models for the second part of the dataset by validating the first set of data models and the determined first fitness function with the second part of the dataset; determining a second fitness function associated with the second part of the dataset based on the generated second set of data models; determining at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function; and outputting the determined at least one best performing data model on a user interface of a user device. ### Claim 16 (depends on 15) The non transitory computer-readable storage medium of claim 15, wherein the one or more parameter ranges includes at least one of a set comprising prediction range, accuracy range, precision range, and recall range. ### Claim 17 (depends on 15) The non transitory computer-readable storage medium of claim 15, wherein the one or more data modelling methods comprises one or more machine learning methods and one or more artificial intelligence methods configured for generating data models. ### Claim 18 (depends on 15) The non transitory computer-readable storage medium of claim 15, wherein the first set of data models represent one or more data patterns inherent in the first part of the dataset and the second set of data models represents one or more data patterns inherent in the second part of the dataset. ### Claim 19 (depends on 15) The non transitory computer-readable storage medium of claim 15, wherein in determining at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function, cause the processor to perform method step: determining at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function using a pareto front method. ### Claim 20 (depends on 15) The non transitory computer-readable storage medium of claim 15, wherein in determining at least one best performing data model among the first set of data models and the second set of data models by comparing the determined first fitness function with the determined second fitness function, cause the processor to perform method step: determining performance of the first set of data models and the second set of data models based on the first fitness function and the second fitness function; and ranking the first set of data models and the second set of data models based on determined performance of the first set of data models and the second set of data models.