Ikaskuntza automatikoa

FLEXIBILE ENTITY RESOLUTION NETWORKS

Resumen de: WO2025175313A1

In various embodiments, a computing system is configured to provide a multi-stage cascade of large language models and stage N neural networks that identifies matching data records within a set of data records and then merges the matching data records. More specifically, the computing system can use a combination of domain-agnostic large language models and downstream neural network classifiers to identify matching data records that would otherwise not be possible with other machine learning or rules-based entity resolution systems. In one example, a computing system receives an entity resolution request. The entity resolution request can indicate a first entity and a second entity. For example, a data steward may provide the entity resolution request to help determine whether the entities are the same or different.

EPISODIC OFF-POLICY EVALUATION OF JOINT ACTIONS FOR A MACHINE-LEARNED POLICY FOR AN ONLINE SYSTEM

Resumen de: US2025265478A1

An off-policy evaluation system performs episodic off-policy evaluations to perform off-policy evaluation (OPE) for multiple, joint episodes. For a single episode, a first machine learning model outputs a propensity for each action for the user and selects a first action for the user from the set of propensities. For a second episode, a second machine learning model outputs a propensity for each action for the user and selects a first action for the user from the set of propensities. The second machine learning model is evaluated by determining an importance weight for the first model and the second model to determine the inverse propensity score of the second machine learning model.

AUTOMATED ARCHITECTURAL SPECIFICATION GENERATION AND HARDWARE IDENTIFICATION

Resumen de: US2025265880A1

A method according to one embodiment includes determining, by a server, a location of a door in an architectural drawing and a room function of a room secured by the door based on an analysis of the architectural drawing, determining, by the server, proper access control hardware to be installed on the door based on the room function, a category of access control hardware, and a predictive machine learning model associated with the category of access control hardware, and generating, by the server, a specification based on the determined proper access control hardware.

Method, System, And Apparatus For Creating Knowledge Graph In Industrial Field

Resumen de: US2025265479A1

Various embodiments of the teachings herein include a method for creating a knowledge graph in the industrial field. An example includes: obtaining unstructured data from a first source in a sub-field of the industrial field, with knowledge annotations; performing machine learning on the unstructured data to generate a first model adapted to extract knowledge; extracting knowledge from second unstructured data provided by the first source based on the first model, without knowledge annotations; obtaining first structured data and first semi-structured data from a second source in a second sub-field; extracting second knowledge from the first structured data; extracting third knowledge from the first semi-structured data; and building a knowledge graph integrating the first and second sub-field based on the first, second, and third knowledge, represented in the form of triples.

SYSTEMS AND METHODS FOR INVENTORY MANAGEMENT AND OPTIMIZATION

Resumen de: US2025265546A1

The present disclosure provides systems and methods that may advantageously apply machine learning to accurately manage and predict inventory variables with future uncertainty. In an aspect, the present disclosure provides a system that can receive an inventory dataset comprising a plurality of inventory variables that indicate at least historical (i) inventory levels, (ii) inventory holding costs, (iii) supplier orders, and/or (iv) lead times over time. The plurality of inventory variables can be characterized by having one or more future uncertainty levels. The system can process the inventory dataset using a trained machine learning model to generate a prediction of the plurality inventory variables. The system can provide the processed in inventory dataset to an optimization algorithm. The optimization algorithm can be used to predict a target inventory level for optimizing an inventory holding cost. The optimization algorithm can comprise one or more constraint conditions.

TECHNIQUES FOR REPORTING CORRELATION METRICS FOR MACHINE LEARNING REPRODUCIBILITY

Resumen de: CN119895449A

Methods, systems, and devices for wireless communication are described. A machine learning server may generate a set of low-dimensional parameters representing training data for the machine learning server, the training data associated with one or more communication environments or one or more channel environments, or a combination thereof. The machine learning server may receive, from one or more devices within a communication environment or a channel environment or both, a set of low-dimensional parameters representing test data associated with the communication environment or the channel environment or both. The machine learning server may generate a reproducibility metric according to a correlation between the set of parameters representing the training data and the set of parameters representing the test data. The machine learning server may send a message indicating the reproducibility metric to the one or more devices, and the one or more devices may perform a communication procedure based on the reproducibility metric.

METHOD AND APPARATUS FOR MONITORING MODEL IN BEAM MANAGEMENT BY USING ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING

Resumen de: EP4604410A1

Provided are a method and apparatus for monitoring a model in beam management by using artificial intelligence and machine learning. The method may include: in relation to a reference signal configured for a terminal, receiving second reference signal resource set configuration information of the reference signal for monitoring an AI/ML model; on the basis of the second reference signal resource set configuration information, measuring signal strength or signal quality for the reference signal; and reporting the performance result of the AI/ML model by comparing a measured value of the reference signal with a predicted value of the reference signal inferred via the AI/ML model.

SYSTEM, METHOD, AND COMPUTER PROGRAM PRODUCT FOR GENERATING A MACHINE LEARNING MODEL BASED ON ANOMALY NODES OF A GRAPH

Resumen de: WO2024081350A1

Provided are systems that include at least one processor to receive a dataset comprising a set of labeled anomaly nodes, a set of unlabeled anomaly nodes, and a set of normal nodes, randomly sample a node to provide a set of randomly sampled nodes, generate a plurality of new nodes based on the set of labeled anomaly nodes and the set of randomly sampled nodes, combine the plurality of new nodes with the set of labeled anomaly nodes to provide a combined set of labeled anomaly nodes, and train a machine learning model based on an embedding of each labeled anomaly node in the combined set of labeled anomaly nodes, a center of the combined set of labeled anomaly nodes in an embedding space, and a center of the set of normal nodes in the embedding space. Methods and computer program products are also disclosed.

SYSTEMS, METHODS, AND GRAPHICAL USER INTERFACES FOR MITIGATING BIAS IN A MACHINE LEARNING-BASED DECISIONING MODEL

Resumen de: US2025259070A1

A system, method, and computer-program product includes obtaining a decisioning dataset comprising a plurality of favorable decisioning records and at least one unfavorable decisioning record; detecting, via a machine learning algorithm, a favorable decisioning record of the plurality of favorable decisioning records that has a vector value closest to a vector value of the unfavorable decisioning record; executing a counterfactual assessment between the favorable decisioning record and the unfavorable decisioning record; generating an explainability artifact based on one or more bias intensity metrics to explain a bias in a machine learning-based decisioning model; and in response to generating the explainability artifact, displaying the explainability artifact in a user interface.

MACHINE-LEARNING-BASED MEAL DETECTION AND SIZE ESTIMATION USING CONTINUOUS GLUCOSE MONITORING (CGM) AND INSULIN DATA

Resumen de: US2025259727A1

Disclosed is a meal detection and meal size estimation machine learning technology. In some embodiments, the techniques entail applying to a trained multioutput neural network model a set of input features, the set of input features representing glucoregulatory management data, insulin on board, and time of day, the trained multioutput neural network model representing multiple fully connected layers and an output layer formed from first and second branches, the first branch providing a meal detection output and the second branch providing a carbohydrate estimation output; receiving from the meal detection output a meal detection indication; and receiving from the carbohydrate estimation output a meal size estimation.

HARD-TO-FIX (HTF) DESIGN RULE CHECK (DRC) VIOLATIONS PREDICTION

Resumen de: US2025258990A1

A method includes: training a machine learning model with a plurality of electronic circuit placement layouts; predicting, by the machine learning model, fix rates of design rule check (DRC) violations of a new electronic circuit placement layout; identifying hard-to-fix (HTF) DRC violations among the DRC violations based on the fix rates of the DRC violations of the new electronic circuit placement layout; and fixing, by an engineering change order (ECO) tool, the DRC violations.

SYSTEMS AND METHODS FOR MANUFACTURING A BATTERY ELECTRODE PLATE

Resumen de: US2025258969A1

The present disclosure relates to systems and methods for manufacturing a battery electrode plate. The system comprises a computing device configured to receive, from the client device, a target process factor among a plurality of process factors associated with manufacturing a battery electrode plate, predict, via a machine-learning model, a change in a characteristic of the battery electrode plate based on a change in a design value of the target process factor, generate information for selecting the target process factor based on predicting the change of the characteristic of the battery electrode plate, and transmit the information to the client device for manufacturing the battery electrode plate.

EMPOWERING RESOURCE-CONSTRAINED IOT EDGE DEVICES: A HYBRID APPROACH FOR EDGE DATA ANALYSIS

Resumen de: US2025259077A1

Methods and systems are provided herein for generating optimized, hybrid machine learning models capable of performing tasks such as classification and inference in IoT environments. The models may be deployed as optimized, task-specific (and/or environment-specific) hardware components (e.g., custom chips to perform the machine learning tasks) or lightweight applications that can operate on resource constrained devices. The hybrid models may comprise hybridization modules that integrate output of one or more machine learning models, according to sets of hyperparameters that are refined according to the task and/or environment/sensor data that will be used by the IoT device.

STABLE CLASSIFICATION BY COMPONENTS FOR INTERPRETABLE MACHINE LEARNING

Resumen de: WO2025168228A1

The present disclosure relates to a stable classification by components (SCBC) data processing architecture, configured to classify input data into one or more classes, comprising: a component detection module configured to compare the input data to a set of detection components, representing data patterns relevant for the classification, and determine a detection probability for each detection component based on the comparison. The SCBC data processing architecture further comprises a probabilistic reasoning module configured to compute one or more class prediction probabilities for the one or more classes based on the determined detection probabilities, a set of class-specific prior probabilities for the determined detection probabilities, and a set of class-specific reasoning probabilities for the determined detection probabilities. Application scenarios include medical and pharmaceutical applications, as well as healthcare in general such as interpretable and secure diagnosis and treatment recommendation systems. Related SCBC data processing system, methods and computer programs are also disclosed, as well as corresponding model training methods and systems.

Sustainable Pipeline Of Pozzolanic Materials

Resumen de: US2025259114A1

In general, in one aspect, embodiments relate to a method of producing a sustainable pipeline of pozzolanic materials that includes gathering unstructured and/or structured data publicly available on a network, identifying analytical data of a pozzolanic material using one or more machine learning models, where the analytical data is present within at least the structured data, extracting the analytical data from the structured data, predicting, using one or more predictive models, one or more performance characteristics of the pozzolanic material based at least in part on the analytical data, to form one or more predicted performance characteristics, comparing the predicted one or more performance characteristics to one or more minimum acceptable performance characteristics, storing the extracted analytical data and the one or more predicted performance characteristics in a database if the one or more performance characteristics meets or exceeds the minimum acceptable performance characteristic, and preparing a cement composition that includes the pozzolanic material if the predicted one or more performance characteristics meets or exceeds the one or more minimum acceptable performance characteristics.

SCORE BASED CERTAINTY ESTIMATION OF PREDICTION

Resumen de: US2025259103A1

The present disclosure describes a patent management system and method for remediating insufficiency of input data for a machine learning system. A prediction to be performed is received from a user input. Relevant input data is determined to perform the prediction. The relevant input data is determined by applying filters based on the prediction to be performed. Prediction is performed by generating a plurality of predicted vectors. A confidence score for the generated plurality of predicted vectors is determined. If the confidence score is less than a predetermined threshold, the prediction is unreliable. The input data is expanded by gathering additional input data. The input data is expanded with the additional input data until the confidence score exceeds the predetermined threshold. A predicted output is generated with the expanded input data. The prediction output and the confidence score are provided for rendering.

APPARATUS AND METHOD FOR EFFICIENTLY OPERATING AI/ML MODEL IN WIRELESS COMMUNICATION SYSTEM

Resumen de: WO2025170089A1

According to various embodiments of the present disclosure, an operation method for a first node in a wireless communication system is provided, the method comprising the steps of: receiving at least one synchronization signal from a second node; receiving control information from the second node; transmitting first communication environment data to the second node; receiving, from the second node, model information related to a first secondary artificial intelligence/machine learning (AI/ML) model based on a first sub-feature set related to the first communication environment data; transmitting, to the second node, second communication environment data changed from the first communication environment data; and receiving, from the second node, model update information for a second secondary AI/ML model, which is based on a second sub-feature set related to the second communication environment data and is changed from the first secondary AI/ML model.

Systems and Methods for Preprocessing Medical Images

Resumen de: US2025259735A1

Systems and methods for preprocessing input images in accordance with embodiments of the invention are disclosed. One embodiment includes a method for performing inference based on input data, the method includes receiving a set of real-valued input images and preprocessing the set of real-valued input images by applying a virtual optical dispersion to the set of real-valued input images to produce a set of real-valued output images. The method further includes predicting, using a machine learning model, an output based on the set of real-valued output images, computing a loss based on the predicted output and a true output, and updating the machine learning model based on the loss.

DATA DIVERSITY VISUALIZATION AND QUANTIFICATION FOR MACHINE LEARNING MODELS

Resumen de: US2025259083A1

Systems and techniques that facilitate data diversity visualization and/or quantification for machine learning models are provided. In various embodiments, a processor can access a first dataset and a second dataset, where a machine learning (ML) model is trained on the first dataset. In various instances, the processor can obtain a first set of latent activations generated by the ML model based on the first dataset, and a second set of latent activations generated by the ML model based on the second dataset. In various aspects, the processor can generate a first set of compressed data points based on the first set of latent activations, and a second set of compressed data points based on the second set of latent activations, via dimensionality reduction. In various instances, a diversity component can compute a diversity score based on the first set of compressed data points and second set of compressed data points.

SEMI-SUPERVISED SYSTEM FOR DOMAIN SPECIFIC SENTIMENT LEARNING

Resumen de: US2025259080A1

Automated computer systems and methods to determine a sentiment of information in digital information or content are disclosed. One aspect includes deriving, by a processor, the digital information from a source; generating, by the processor, a domain-specific machine learning sentiment score, based on the digital information, by one model of at least two machine learning models; autonomously mapping, by the processor, a non-domain specific knowledge graph of associations between elements in a set of digital contextual information; receiving, by the processor, sentiment graphs, each sentiment graph defining a sentiment; generating, by the processor, a graph sentiment score based on the non-domain specific knowledge graph and the sentiment graphs; generating, by the processor, a final sentiment score based on the graph sentiment score and the domain-specific machine learning sentiment score; and determining the sentiment of the information in the digital information or content via the final sentiment score.

AI DISCOVERY

Resumen de: WO2025166404A1

This disclosure relates generally to detecting artificial intelligence (AI) implementation in a software application comprising one or more application packages (APs). One or more processors extract one or more AP strings from the software application, which each represent an AP; and create a prompt for a machine learning model, trained to generate output text, comprising the one or more AP strings, the prompt representing instructions to provide a classification and provide functionality information of each of the one or more APs, the classification being AI relevant or non-AI relevant and the functionality information describing a functionality of the respective AP. The one or more processors then evaluate the machine learning model on the prompt to generate output text corresponding to the classification and the functionality information of each of the one or more APs; and generate a report of the AI implementation based on the output text.

OBTAINING INFERENCES TO PERFORM ACCESS REQUESTS AT A NON-RELATIONAL DATABASE SYSTEM

Resumen de: US2025258821A1

Inferences may be obtained to handle access requests at a non-relational database system. An access request may be received at a non-relational database system. The non-relational database system may determine that the access request uses a machine learning model to complete the access request. The non-relational database system may cause an inference to be generated using data items for the access request as input to the machine learning model. The access request may be completed using the generated inference.

SPECTRAL ANALYSIS, MACHINE LEARNING, AND FRAC SCORE ASSIGNMENT TO ACOUSTIC SIGNATURES OF FRACKING EVENTS

Resumen de: US2025258311A1

System, method, and apparatus for classifying fracture quantity and quality of fracturing operation activities during hydraulic fracturing operations, the system comprising: a sensor coupled to a fracking wellhead, circulating fluid line, or standpipe of a well and configured to convert acoustic vibrations in fracking fluid in the fracking wellhead into an electrical signal; a memory configured to store the electrical signal; a converter configured to access the electrical signal from the memory and convert the electrical signal in a window of time into a current frequency domain spectrum; a machine-learning system configured to classify the current frequency domain spectrum, the machine-learning system having been trained on previous frequency domain spectra measured during previous hydraulic fracturing operations and previously classified by the machine-learning system; and a user interface configured to return a classification of the current frequency domain spectrum to an operator of the fracking wellhead.

MALICIOUS UNIFORM RESOURCE LOCATOR (URL) DETECTION

Resumen de: US2025258917A1

Apparatuses, systems, and techniques for classifying a candidate uniform resource locator (URL) as a malicious URL using a machine learning (ML) detection system. An integrated circuit is coupled to physical memory of a host device via a host interface. The integrated circuit hosts a hardware-accelerated security service that obtains a snapshot of data stored in the physical memory and extracts a set of features from the snapshot. The security service classifies the candidate URL as a malicious URL using the set of features and outputs an indication of the malicious URL.

IDENTIFICATION OF NETWORK EVENTS REPRESENTING A NETWORK SECURITY THREAT

Nº publicación: WO2025163013A1 07/08/2025

Solicitante:

VOCALINK LTD [GB]
VOCALINK LIMITED

Resumen de: WO2025163013A1

A computer-implemented method is provided for training a machine learning model to identify one or more network events associated with a network and representing a network security threat. The method comprises: a) obtaining a first dataset comprising data representative of a plurality of network events in a first network; b) obtaining a second dataset comprising data representative of a plurality of network events in a second network; c) performing covariate shift analysis on the first dataset and the second dataset to identify and classify a plurality of differences between the first dataset and the second dataset; d) performing domain adaptation on the first dataset, based on a classified difference, to generate a training dataset; e) training a machine learning model using the training dataset to produce a trained threat detection model.