International Business Machines Corporation

Kyndryl, Inc.

Joseph N Kozhaya

Corville O Allen

Shikhar Kwatra

Andrew Ronald Freed

Ryan R Anderson

Christopher Michael Madison

Lakshminarayanan Krishnamurthy

Clyde T Foster, Ii

Swaminathan Chandrasekaran

Dwi Sianto Mansjur

Sridhar Sudarsan

Anne Elizabeth Gattiker

Anita Govindjee

Javier Torres

Jhansi R Kolla

Sai Prathyusha Peddi

Mehul H Darji

Dana Lea Price

Bernadette Alexia Carter

Phillip John Restle

Rahul Ghosh

Sourav Mazumder

Phaedra Boinodiris

Pritesh Patel

Anantha S Rao

Valentina Salapura

Craig M Trim

Zhuo Li

Charles Jay Alpert

Chin Ngai Sze

John Maxwell Cohn

Rama Kalyani T Akkiraju

David Osmond Melville

Haifeng Qian

Ying Zhou

Dario Gil

Andrew Dwight Dingsor

Neil Sahota

Joseph J Palumbo

Veena S Pureswaran

Michael T Peters

Teppei Tsurumi

Robert A Proctor

Douglas Alan Williams

John D Wolpert

Craig Cook

Adrian B Jordan

David S Richardson

Kristin M Arrowood

Aaron K Baughman

Paul Gerard Villarrubia

Christopher J Berry

Michael H Wood

Adam Lee Griffin

Ivan L Wemple

Patrick H Buffet

Ciaran J Brennan

Laura Janet Rodriguez

Jenna Reinen

Erich C Schanzenbach

Trenton J Johnson

David Wen-Hao Shan

Alissa J Hartenbaum

Daniel Richard Menard

Patrick M Ryan

Amanda Christine Venton

Rachel Ostrand

Karl David McCormick, Ii

John Henry Welborn, Jr

William A Beason

Elizabeth Spingola

Marcos Araque Fiallos

Paul Samuel Schweigert

John F Gbruoski

Mehrdad Ashrafian

Vincent J Dowling

Susan R Sanicky

Lu′Ay Bakir

Lu'ay Bakir

Paul D Montane

Vicki L Flaherty

Jennifer K Paylor

An approach for intelligent optimization of machine learning models for a target environment may be provided herein. The approach may include extracting metadata from a training pipeline for a first machine learning model that has been configured to operate within a first computing environment. The approach may also include mapping the extracted metadata to one or more constraints associated with a second machine learning model that has been configured to operate within a second computing environment. The approach may also include training the second machine learning model configured to the second computing environment, with the dataset that was used to train the first machine learning model, based on the mapped constraints. The approach may also include comparing performance metrics of the first machine learning model to corresponding metrics of the now trained second machine learning model.

Intelligently optimized machine learning models

According to aspects of the present disclosure, systems, methods and computer program products can be provided for dynamic workload orchestration based on data complexity. Methods, computer program products and/or systems are provided for dynamic workload orchestration that perform the following operations: (i) receiving a workload for orchestration; (ii) computing complexity scores for respective portions of the workload, where the complexity scores are computed based at least on parameters describing data associated with the portions of the workload; and (iii) using an orchestration engine to assign the portions of the workload to corresponding compute resources, based on their respective complexity scores.

Dynamic data driven orchestration of workloads

In an approach to monitoring and querying autobiographical events, one or more computer processors receive data from one or more sensors associated with a user. Based on the received data from the one or more sensors, one or more computer processors detect an autobiographical event associated with the user. One or more computer processors convert the received data from the one or more sensors into one or more observations associated with the autobiographical event. One or more computer processors convert the one or more observations into one or more questions associated with the one or more observations. One or more computer processors query the user with the one or more questions. One or more computer processors receive at least one response to the one or more questions from the user.

Monitoring and querying autobiographical events

A computer-implemented method for adapting a deep learning model to a local environment includes collecting training data; training a common deep learning model using the collected training data; customizing the deep learning model based on characteristics specific to one of a plurality of local devices utilizing transfer learning; and deploying the customized deep learning model to the one of the plurality of local devices. The customized deep learning model may be trained across a hierarchy comprising: a central common model trained on data from all devices in the plurality of devices; a location based model trained on a first set of data from devices in the plurality of devices within a location proximity; and a device based model trained on a data from one device in the plurality of devices.

Scalable and dynamic transfer learning mechanism

An asset stored on a storage device of a computer system is detected to be omitted from an asset-specific access policy. Using a utilization history of the asset, a criticality of the asset is scored, generating a first criticality score for the asset. Using the first criticality score, the asset is classified into a policy category. A selected access policy selected from a set of access policies assigned to a set of assets in the policy category is applied to the asset, the selected access policy specifying an access restriction of the asset.

Similarity based digital asset management

A processor may receive data regarding a context for a first dialog turn. The processor may monitor a voice input from a user for the first dialog turn. The processor may detect a first pause in the voice input, the first pause having a duration that satisfies a time threshold. The processor may receive, based on the first pause, first voice input data. The processor may analyze the first voice input data. The processor may determine that additional time is recommended for the voice input to be provided by the user.

Dynamic voice input detection for conversation assistants

Mechanisms are provided for implementing a virtual corpus engine that receives an inquiry to be processed and analyzes the inquiry to extract one or more features of the inquiry. The virtual corpus engine selects a weight matrix associated with a virtual corpus based on the extracted one or more features of the inquiry. The virtual corpus comprises a plurality of actual corpora of information. The weight matrix comprises a separate weight value for each actual corpus in the plurality of actual corpora. The virtual corpus engine processes the inquiry using a set of selected actual corpora selected from the plurality of actual corpora based on the weight values in the weight matrix and receives results of the processing of the inquiry using the set of selected actual corpora. The virtual corpus engine outputs the results of the processing of the inquiry.

Cognitive system virtual corpus training and utilization

Clustering a set of natural language queries NLQs based on a set of significant events retrieved from a corpus stored in a computer system is described. A set of NLQs is used by a search engine for searching a selected corpus to retrieve respective sets of significant events. The set of NLQs is clustered into a plurality of NLQ clusters according to a number of common significant events being returned by the search engine for respective members of an NLQ cluster.

Clustering a set of natural language queries based on significant events

The present invention provides a method, computer program product, and system of performing personality profile automated measurements. In an embodiment, the method, computer program product, and system includes receiving, by a computer system, gaming history data associated with a gamer profile, assigning, by the computer system, the gamer profile to a first cluster of gamers, recording, by the computer system, game play data associated with the gamer profile, and based on the game play data associated with the gamer profile and the gaming history data associated with the gamer profile, reassigning the gamer profile into a second cluster of gamers.

Reassigning gamer clusters based on engagement

Embodiments relate to an intelligent computer platform to identify a lexical answer type (LAT), a first concept relevant to the received request and a second concept related to the identified first concept. The LAT, together with the first and second concepts are utilized to create a first and second cluster. Documents are selectively populated into the clusters. The clusters are subject to sorting based on a relevancy protocol.

Dynamic query processing and document retrieval

Microservices are predictively deployed on edge devices in a network. An application is run on a client device, the application comprising a set of microservices runnable on any edge device in a set of two or more edge devices. A state of the client device at a first time is determined, the state including one or more microservices currently being run for the client device, and for each microservice currently being run, an edge device running the microservice. One or more microservices that are likely to be run at a second time subsequent to the first time and a location of the client device at the second time are predicted. Based on the predicted location, a next edge device in the set of edge devices is determined for running the one or more microservices predicted to be run at the second time.

Intelligent dynamic communication handoff for mobile applications

A method identifies and removes bias from a machine learning model. A user/computer inputs a plurality of input training data into a machine learning system to generate an output of labeled output data. The user/computer evaluates the labeled output data according to a consistency metric to associate the labeled output data with a corresponding consistency assessment. The user/computer selects each labeled output data having a consistency assessment indicating a consistency assessment that is greater than a predetermined threshold to form a labeled output data subset, and then creates additional labeling for the labeled output data subset. The user/computer utilizes the additional labeling to distinguish each labeled training data from labeled output data subset as being mislabeled and biased, and then adjusts the learning machine based on the labeled output data subset being mislabeled and biased.

Detecting and reducing bias in machine learning models

A method, system, and computer program product provide navigation guidance around various object types in a vicinity of a guidance assistance device for a user by analyzing received data regarding an environment around the guidance assistance device to identify one or more entities Ei and by applying an artificial intelligence machine learning analysis to group the one or more entities Ei into corresponding categories Cj and to determine a minimum spacing distance Dmin for each of the one or more entities Ei, wherein the minimum spacing distance Dmin is a minimum distance between the guidance assistance device and the entity Ei based on the categorization Ci specified by the user, and then providing feedback to the user when any of the one or more entities Ei is within than minimum spacing distance Dmin corresponding to said entity.

Assistive mechanism via edge device personalization

A method and system for delivering a package is provided. The method includes receiving online order data comprising details associated with a product being ordered by a user. The online order data and environmental data associated with delivery of the product is analyzed and a resulting a travel route for delivering the product to a specified location of the user is generated. A delivery vehicle for delivering the product to the specified location of the user is selected and directed to a storage location of the product for loading. The delivery vehicle is directed from the storage location to the specified location of the user via the travel route. Delivery actions associated with the product with respect to the specified location are determined based on data indicating current environmental conditions. The delivery actions are executed.

Intelligent vehicle delivery

A method, system and computer-usable medium are disclosed for improving likelihood of user to purchase a product or service. Interactions of a person, not necessarily the user, are monitored related to the product or service, along with actions by a business related to the person and the product or service. A trained Generative Adversarial Network model is applied to the monitored interactions to form recommend actions that the business should take to achieve ameliorative actions by the user. A reward feedback iterative adjustment of the GAN model is used to facilitate purchase of the product or service by the user.

System and method for generating preferred ameliorative actions using generative adversarial networks

Growing community of inventors

Morrisville, NC, United States of America

Joseph N Kozhaya