Patents Assigned to Pulselight Holdings, Inc.
-
Patent number: 12217875Abstract: A method for generating synthetic training records for use in training a model to predict low-incidence events. A synthetic training record is generated from a minority-class training record by substituting a different value for a feature in the minority-class training record, where the probability of the different value occurring in the minority-class training record exceeds a probability threshold. Also disclosed are a non-transitory storage medium comprising minority-class training records and synthetic training records and a method of training a machine-leaning model using training records augmented with synthetic training records. An exemplary synthetic training records is a synthetic medical record for use in training a model to predict drug overdoses.Type: GrantFiled: October 31, 2022Date of Patent: February 4, 2025Assignee: Pulselight Holdings, Inc.Inventors: Jonathan Mugan, Mallika Thanky
-
Publication number: 20240296355Abstract: A system of machine learning (“ML”) models for making actionable predictions regarding low-incidence events, including a generative ML model that produces synthetic minority-class records to form an augmented training data set, a predictive ML model that has been trained on the augmented training data set, a certainty ML model that produces a certainty estimate, and an explanatory model that produces an explanation. A method for producing actionable predictions of a low-incidence event by applying ML models to imbalanced class data by producing a prediction by a predictive ML model that has been trained on a data set comprising synthetic minority-class data records produced by a generative ML model, and producing a certainty estimate and an explanation. At least one of the certainty estimate or explanation determines an effective or appropriate response to the prediction. The low-incidence event may comprise risk of opioid use disorder.Type: ApplicationFiled: May 6, 2024Publication date: September 5, 2024Applicant: Pulselight Holdings, Inc.Inventors: Jonathan Mugan, Mallika Thanky
-
Publication number: 20240265034Abstract: Disclosed are a method and system for propagating data changes in a hierarchy of dataset models in which each dataset model comprises an analytic and one or more parent datasets, including a primordial dataset. The analytic is executed to instantiate a first instance of the data model. After a change in a primordial dataset, each instance of a dataset model that descends from the primordial dataset is invalidated, and the analytic is re-executed to create a second instance of the data model. Analytical results may be displayed. The first dataset model may include a metric in which the definition of the metric comprises metadata of the dataset model. Metric values may be stored in a first cache, re-computed on a new instance of the dataset model, and stored in a second cache.Type: ApplicationFiled: April 2, 2024Publication date: August 8, 2024Applicant: Pulselight Holdings, Inc.Inventors: James SNYDER, Stuart JARRIEL, Joseph Raphael Dente
-
Patent number: 11977991Abstract: A system of machine learning (“ML”) models for making actionable predictions regarding low-incidence events, including a generative ML model that produces synthetic minority-class records to form an augmented training data set, a predictive ML model that has been trained on the augmented training data set, a certainty ML model that produces a certainty estimate, and an explanatory model that produces an explanation. A method for producing actionable predictions of a low-incidence event by applying ML models to imbalanced class data by producing a prediction by a predictive ML model that has been trained on a data set comprising synthetic minority-class data records produced by a generative ML model, and producing a certainty estimate and an explanation. At least one of the certainty estimate or explanation determines an effective or appropriate response to the prediction. The low-incidence event may comprise risk of opioid use disorder.Type: GrantFiled: July 24, 2020Date of Patent: May 7, 2024Assignee: Pulselight Holdings, Inc.Inventors: Jonathan Mugan, Mallika Thanky
-
Patent number: 11947567Abstract: Disclosed are a method and system for propagating data changes in a hierarchy of dataset models in which each dataset model comprises an analytic and one or more parent datasets, including a primordial dataset. The analytic is executed to instantiate a first instance of the data model. After a change in a primordial dataset, each instance of a dataset model that descends from the primordial dataset is invalidated, and the analytic is re-executed to create a second instance of the data model. Analytical results may be displayed. The first dataset model may include a metric in which the definition of the metric comprises metadata of the dataset model. Metric values may be stored in a first cache, re-computed on a new instance of the dataset model, and stored in a second cache.Type: GrantFiled: November 15, 2021Date of Patent: April 2, 2024Assignee: Pulselight Holdings, Inc.Inventors: James Snyder, Stuart Jarriel, Joseph Raphael Dente
-
Publication number: 20230409638Abstract: Observable data points are collected and organized into a link-oriented data set comprising nodes and links. Information is abstracted for use in link analysis by generating links between the collected data points, including deriving links and inducing links. A link can be induced by linking together a pair of nodes that satisfy a distance function. Exemplary distance functions that can be used to induce links include geospatial proximity, attribute nearness, and name similarity. Paths can be identified between selected nodes of interest through a dataset operation, and nodes and/or links can be selectively included or excluded from the data set operation. The dataset can be augmented with pedigree information or one or more association nodes. Link information, including a trajectory and a connected path that selectively produces or excludes one or more intermediate nodes, can be displayed and/or produced in a specified format.Type: ApplicationFiled: August 30, 2023Publication date: December 21, 2023Applicant: Pulselight Holdings, Inc.Inventors: Jim SNYDER, Joon Hao CHUAH, Joe DENTE, Travis HARTWELL, Morgan HOLLINGER, John THELE, Jimmy WAN, Robert WILLIAMS, Robby MORGAN
-
Patent number: 11762909Abstract: Observable data points are collected and organized into a link-oriented data set comprising nodes and links. Information is abstracted for use in link analysis by generating links between the collected data points, including deriving links and inducing links. A link can be induced by linking together a pair of nodes that satisfy a distance function. Exemplary distance functions that can be used to induce links include geospatial proximity, attribute nearness, and name similarity. Paths can be identified between selected nodes of interest through a dataset operation, and nodes and/or links can be selectively included or excluded from the data set operation. The dataset can be augmented with pedigree information or one or more association nodes. Link information, including a trajectory and a connected path that selectively produces or excludes one or more intermediate nodes, can be displayed and/or produced in a specified format.Type: GrantFiled: July 6, 2021Date of Patent: September 19, 2023Assignee: Pulselight Holdings, Inc.Inventors: Jim Snyder, Joon Hao Chuah, Joe Dente, Travis Hartwell, Morgan Hollinger, John Thele, Jimmy Wan, Robert Williams, Robby Morgan
-
Patent number: 11488723Abstract: A method for generating synthetic training records for use in training a model to predict low-incidence events. A synthetic training record is generated from a minority-class training record by substituting a different value for a feature in the minority-class training record, where the probability of the different value occurring in the minority-class training record exceeds a probability threshold. Also disclosed are a non-transitory storage medium comprising minority-class training records and synthetic training records and a method of training a machine-leaning model using training records augmented with synthetic training records. An exemplary synthetic training records is a synthetic medical record for use in training a model to predict drug overdoses.Type: GrantFiled: June 3, 2019Date of Patent: November 1, 2022Assignee: Pulselight Holdings, Inc.Inventors: Jonathan Mugan, Mallika Thanky
-
Publication number: 20220269707Abstract: A recurrent neural network (RNN) method implemented on a computer system is used to produce summaries of unstructured text generated by multiple networks of individuals interacting over time by encoding the unstructured text into intermediate representations and decoding the intermediate representations into summaries of each network. Parameter data for the RNN is obtained by using multiple different versions of the same source texts to train the computer system. The method and computer system can be used to identify which of the networks match a query by determining which network generates the query with low or lowest cost.Type: ApplicationFiled: February 28, 2022Publication date: August 25, 2022Applicant: PULSELIGHT HOLDINGS, INC.Inventors: JONATHAN WILLIAM MUGAN, LAURA HITT, JIMMIE GOODE, RUSS GREGORY, YUAN QU
-
Publication number: 20220075800Abstract: Disclosed are a method and system for propagating data changes in a hierarchy of dataset models in which each dataset model comprises an analytic and one or more parent datasets, including a primordial dataset. The analytic is executed to instantiate a first instance of the data model. After a change in a primordial dataset, each instance of a dataset model that descends from the primordial dataset is invalidated, and the analytic is re-executed to create a second instance of the data model. Analytical results may be displayed. The first dataset model may include a metric in which the definition of the metric comprises metadata of the dataset model. Metric values may be stored in a first cache, re-computed on a new instance of the dataset model, and stored in a second cache.Type: ApplicationFiled: November 15, 2021Publication date: March 10, 2022Applicant: Pulselight Holdings, Inc.Inventors: James SNYDER, Stuart JARRIEL, Joseph Raphael Dente
-
Patent number: 11263250Abstract: A recurrent neural network (RNN) method implemented on a computer system is used to produce summaries of unstructured text generated by multiple networks of individuals interacting over time by encoding the unstructured text into intermediate representations and decoding the intermediate representations into summaries of each network. Parameter data for the RNN is obtained by using multiple different versions of the same source texts to train the computer system. The method and computer system can be used to identify which of the networks match a query by determining which network generates the query with low or lowest cost.Type: GrantFiled: October 14, 2019Date of Patent: March 1, 2022Assignee: Pulselight Holdings, Inc.Inventors: Jonathan William Mugan, Laura Hitt, Jimmie Goode, Russ Gregory, Yuan Qu
-
Patent number: 11188827Abstract: A computer-implemented method of creating synthetic training data records for machine learning, comprising a source computer system operatively connected to a separate target computer system, where the target system comprises real data records, the contents of which are not known to the source computer system. The source computer system comprises the generator component of a generative adversarial network (GAN) and the target computer system comprises a discriminator component of the GAN. The source computer system generates one or more synthetic training records using a computational process that does not reveal contents of the real data records. Also disclosed is a method of training a machine-learning model using the one or more synthetic training records.Type: GrantFiled: June 3, 2019Date of Patent: November 30, 2021Assignee: Pulselight Holdings, Inc.Inventor: Jonathan Mugan
-
Patent number: 11176175Abstract: Disclosed are a method and system for propagating data changes in a hierarchy of dataset models in which each dataset model comprises an analytic and one or more parent datasets, including a primordial dataset. The analytic is executed to instantiate a first instance of the data model. After a change in a primordial dataset, each instance of a dataset model that descends from the primordial dataset is invalidated, and the analytic is re-executed to create a second instance of the data model. Analytical results may be displayed. The first dataset model may include a metric in which the definition of the metric comprises metadata of the dataset model. Metric values may be stored in a first cache, re-computed on a new instance of the dataset model, and stored in a second cache.Type: GrantFiled: July 31, 2015Date of Patent: November 16, 2021Assignee: Pulselight Holdings, Inc.Inventors: James Snyder, Stuart Jarriel, Joseph Raphael Dente
-
Publication number: 20210342398Abstract: Observable data points are collected and organized into a link-oriented data set comprising nodes and links. Information is abstracted for use in link analysis by generating links between the collected data points, including deriving links and inducing links. A link can be induced by linking together a pair of nodes that satisfy a distance function. Exemplary distance functions that can be used to induce links include geospatial proximity, attribute nearness, and name similarity. Paths can be identified between selected nodes of interest through a dataset operation, and nodes and/or links can be selectively included or excluded from the data set operation. The dataset can be augmented with pedigree information or one or more association nodes. Link information, including a trajectory and a connected path that selectively produces or excludes one or more intermediate nodes, can be displayed and/or produced in a specified format.Type: ApplicationFiled: July 6, 2021Publication date: November 4, 2021Applicant: Pulselight Holdings, Inc.Inventors: Jim SNYDER, Joon Hao CHUAH, Joe DENTE, Travis HARTWELL, Morgan HOLLINGER, John THELE, Jimmy WAN, Robert WILLIAMS, Robby MORGAN
-
Patent number: 11055350Abstract: Observable data points are collected and organized into a link-oriented data set including nodes and links. Information is abstracted for use in link analysis by generating links between the collected data points, including deriving links and inducing links. A link can be induced by linking together a pair of nodes that satisfy a distance function. Exemplary distance functions that can be used to induce links include geospatial proximity, attribute nearness, and name similarity. Paths can be identified between selected nodes of interest through a dataset operation, and nodes and/or links can be selectively included or excluded from the data set operation. The dataset can be augmented with pedigree information or one or more association nodes. Link information, including a trajectory and a connected path that selectively produces or excludes one or more intermediate nodes, can be displayed and/or produced in a specified format.Type: GrantFiled: May 25, 2018Date of Patent: July 6, 2021Assignee: Pulselight Holdings, Inc.Inventors: Jim Snyder, Joon Hao Chuah, Joe Dente, Travis Hartwell, Morgan Hollinger, John Thele, Jimmy Wan, Robert Williams, Robby Morgan
-
Patent number: 10445356Abstract: A recurrent neural network (RNN) method implemented on a computer system is used to produce summaries of unstructured text generated by multiple networks of individuals interacting over time by encoding the unstructured text into intermediate representations and decoding the intermediate representations into summaries of each network. Parameter data for the RNN is obtained by using multiple different versions of the same source texts to train the computer system. The method and computer system can be used to identify which of the networks match a query by determining which network generates the query with low or lowest cost.Type: GrantFiled: June 23, 2017Date of Patent: October 15, 2019Assignee: Pulselight Holdings, Inc.Inventors: Jonathan William Mugan, Laura Hitt, Jimmie Goode, Russ Gregory, Yuan Qu
-
Publication number: 20180276311Abstract: Observable data points are collected and organized into a link-oriented data set comprising nodes and links. Information is abstracted for use in link analysis by generating links between the collected data points, including deriving links and inducing links. A link can be induced by linking together a pair of nodes that satisfy a distance function. Exemplary distance functions that can be used to induce links include geospatial proximity, attribute nearness, and name similarity. Paths can be identified between selected nodes of interest through a dataset operation, and nodes and/or links can be selectively included or excluded from the data set operation. The dataset can be augmented with pedigree information or one or more association nodes. Link information, including a trajectory and a connected path that selectively produces or excludes one or more intermediate nodes, can be displayed and/or produced in a specified format.Type: ApplicationFiled: May 25, 2018Publication date: September 27, 2018Applicant: Pulselight Holdings, Inc.Inventors: Jim SNYDER, Joon Hao CHUAH, Joe DENTE, Travis HARTWELL, Morgan HOLLINGER, John THELE, Jimmy WAN, Robert WILLIAMS, Robby MORGAN