Alerta

METHOD, DEVICE, AND SYSTEM FOR DETECTING ABNORMALITY OF SEMICONDUCTOR EQUIPMENT

Resumen de: US2025348760A1

Provided is a method, performed by a computing system, of detecting an abnormality, the method including generating, a plurality of training data sets from respective history data of the plurality of pieces of equipment, generating, based on the plurality of training data sets, a plurality of machine learning models respectively corresponding to the plurality of pieces of equipment, determining, based on a plurality of feature importance sets respectively corresponding to the plurality of machine learning models, a plurality of inefficiency indices respectively corresponding to the plurality of pieces of equipment, and identifying, based on the plurality of inefficiency indices, at least one piece of abnormal equipment among the plurality of pieces of equipment.

METHODS AND SYSTEMS FOR DETERMINING DENTAL CARIES

Resumen de: US2025349433A1

A sequencing module configured to provide metatranscriptomic reads from an oral sample from a subject; metatranscriptomic reads from the sequencing of oral sample and identify and cluster microbes identified in the oral sample into taxon clusters (TCs) using the metatranscriptomic reads mapped to a metagenomic library; generate TC-specific orthogroups for each of the TCs via protein clustering; determine KEGG orthology for each of the TC-specific orthogroups, or genes directly; generate phylogenomic functional categories (PGFCs) from grouping of gene expression counts by the KEGG modules for each of the TCs; retain the PGFCs having an MCR above an MCR threshold to obtain input data; and identify or predict, using a classifier model including variables selected by a feature selection machine learning algorithm, dental caries in said subject based on the input data.

SYSTEMS AND METHODS FOR ANALYZING AND MITIGATING COMMUNITY-ASSOCIATED RISKS

Resumen de: US2025348967A1

A computer system for analyzing and mitigating risks associated with a building is provided. The computer system is configured to: (i) receive environment data from the at least one sensor; (ii) receive building data from the at least one database; (iii) utilize a trained machine learning model to determine at least one potential risk associated with the building based upon the environment data and the building data; (iv) generate a building risk profile that includes the at least one potential risk associated with the building; and/or (v) generate a risk mitigation output based upon at least one of the building risk profile and the at least one potential risk, wherein the risk mitigation output includes at least one of a risk alert, a risk mitigation recommendation, and risk mitigation instructions. Computer systems for analyzing and mitigation risks associated with a city, a user, and an event are also provided.

INPUT SELECTION AWARE MONITORING FOR ENHANCED MACHINE LEARNING BASED POSITIONING

Resumen de: WO2025233559A1

A method for a first apparatus, the method comprising: receiving (701), from the second apparatus, a request to initiate obtaining at least one matching factor, wherein the at least one matching factor indicates a degree matching between inputs used during a machine learning training phase and inputs used during a machine learning inference phase; obtaining (703) the at least one matching factor based on the request to initiate obtaining the at least one matching factor; performing (709) at least a machine learning model switching based on a monitoring decision request (707) from a second apparatus, the monitoring decision request based at least on: the obtained at least one matching factor; and at least one channel characteristic.

UNIFORM ANALYSIS OF MULTILINGUAL RECORDS VIA KNOWLEDGE GRAPH TRANSLATION AND TRANSFER LEARNING

Resumen de: WO2025233668A1

A computer-implemented method includes receiving input text documents in a majority language and training a plurality of models using the input text documents. Training the plurality of models using the input text documents includes conjunctively training a text to knowledge graph model in the majority language and a text to knowledge graph model in the minority language and conjunctively training a text translation model and a knowledge graph translation model, wherein one or more weights from the text to text translation model is shared with the knowledge graph translation model. The method has applications including, but not limited to, use cases in machine learning and medicine / healthcare, e.g., performing uniform analysis of multilingual records of a patient via knowledge graph translation and transfer learning, and improving text to knowledge graph extraction and knowledge graph translation for medical records in multiple languages, which can help optimize decision making.

ROBOTIC MEDICAL SYSTEM WITH MACHINE LEARNING PREDICTED PROCEDURE OUTCOMES FOR SCENARIOS

Resumen de: WO2025235630A1

Machine learning predicted outcomes for scenarios of a procedure performed with a robotic medical system is provided. A system can identify state information related to a medical procedure to be performed by a robotic medical system. The system can generate, using the state information and with one or more models trained by machine learning, scenarios of operation of the robotic medical system to perform the medical procedure. The system can determine, using the one or more models, performance metrics for the scenarios. The system can select, based on a comparison of the performance metrics for the scenarios, a scenario of the scenarios for operation. The system can provide an indication of the selected scenario to cause the robotic medical system to perform at least a portion of the medical procedure in accordance with the selected scenario.

INPUT SELECTION AWARE MONITORING FOR ENHANCED MACHINE LEARNING BASED POSITIONING

Resumen de: WO2025233561A1

A method for a first apparatus, the method comprising: obtaining (753) at least one matching factor, wherein the at least one matching factor indicates a degree matching between inputs used during a machine learning training phase and inputs used during a machine learning inference phase; performing (759) at least a machine learning model switching based on a monitoring decision request from a second apparatus, the monitoring decision request based at least on: the obtained at least one matching factor; and at least one channel characteristic.

NODAL SELF-DETERMINATION BASED ON COMPUTATIONAL COST

Resumen de: WO2025234984A1

A node within a machine learning model may self-determine its own nodal actions by calculating predicted computational costs for nodal actions in a group of nodal actions performable by the node. The node selects a nodal action from the group based on one or more comparisons among the predicted computational costs. The node then performs the nodal action that was selected based on the one or more comparisons of the predicted computational costs. A node manager within a machine learning model may manage a node within the machine learning model by calculating predicted computational costs for nodal commands in a group of nodal commands performable on the node. The node manager selects a nodal command to be performed on the node from the group based on one or more comparisons among the predicted computational costs. The node manager then causes performance of the selected nodal command on the node.

METHOD FOR IMPROVING ACCURACY OF MACHINE LEARNING MODELS

Resumen de: WO2025233706A1

Broadly speaking, embodiments of the present techniques provide a method for training neuro-symbolic ML models through partial label learning, to improve the accuracy of the predictions of the ML models. Advantageously, the present techniques provide a training method which improves how well the ML model is trained, by reducing the training problem to that of partial label learning. In other words, the present techniques train the neural module through partial label learning, rather than conventional multi-class learning techniques.

SYSTEM FOR VISUALIZING CARBON EMISSIONS OF MACHINE LEARNING MODEL

Resumen de: WO2025234546A1

The present invention discloses a system for visualizing carbon emissions. More specifically, the present invention relates to a system for calculating and visualizing carbon emissions generated by applying a machine learning model to various technical fields. According to an embodiment of the present invention, a developer accesses a system online to obtain information on carbon emissions for various ML models, selects a model to be reviewed among a plurality of registered ML models, and uploads prepared query data, and the system converts and inputs the query data into the ML model, and provides, in a graph form, carbon emissions for each inference process generated by performing analysis, so that the developer can easily check the carbon emissions of the ML model being developed thereby.

NODAL SELF-DETERMINATION BASED ON FREE ENERGY

Resumen de: WO2025234983A1

A method includes calculating, by a node within a machine learning model, free energy measures corresponding to all nodal actions in a predetermined plurality of nodal actions that are performable by the node within the machine learning model during an upcoming timestep. A nodal action is selected to be performed by the node during the upcoming timestep based on comparisons among the free energy measures. The node performs the selected nodal action during the upcoming timestep.

Input selection aware monitoring for enhanced machine learning based positioning

Resumen de: GB2640955A

A method performed between a user equipment (UE) or gNodeB (gNB) and a location management function (LMF). The UE/gNB obtains at least one matching factor, which indicates a degree matching between inputs used during a machine learning (ML) training phase and inputs used during a ML inference phase. The UE/gNB performs ML model switching based on a monitoring decision request from the LMF, the monitoring decision request is based on the obtained matching factor and at least one channel characteristic. The monitoring decision request may be further based on a positioning accuracy parameter. The inputs may be TRP measurements for a location management service or positioning determination. The LMF may request the UE/gNB to initiate obtaining the matching factor, and the request may include a threshold matching factor corresponding to the at least one channel characteristic and/or the positioning accuracy parameter.

METHOD, SYSTEM, AND COMPUTER PROGRAM PRODUCT FOR IMPROVING MACHINE LEARNING MODELS

Resumen de: US2025165874A1

Methods, systems, and computer program products are provided for improving machine learning models which include receiving a data set including data records; inputting the data set to a pre-trained first machine learning model to generate first embeddings; inputting the first embeddings to a second machine learning model to generate second embeddings in a user-specific embedding space; inputting the plurality of second embeddings to a third machine learning model to extract feature data associated with a feature; inputting an output from a machine learning system and the feature data to a fourth machine learning model to generate a relevance score for each entity; determining a subset of entities based on the relevance score; communicating a feedback request to a user; receiving feedback data from the user; and training at least one of the models based on the feedback data.

SYSTEMS AND METHODS FOR TEXT CLASSIFICATION

Resumen de: EP4647949A1

A computer-implemented method for text classification may include prompting, by a computing device, a large language model (LLM) to extract a set of explanatory statements for a training dataset, wherein each explanatory statement describes a difference between a grouping of samples within the training dataset. The method may also include prompting, by the computing device, the LLM to generate a set of queries based on the set of explanatory statements, wherein each query evaluates the difference described by an explanatory statement. The method may include training, by the computing device, a machine-learning (ML) model using the set of queries as features. Furthermore, the method may include classifying, by the computing device using the ML model, an unknown sample as a security threat. Finally, the method may include performing, by the computing device, a security action to mitigate the security threat. Various other methods, systems, and computer-readable media are also disclosed.

Training generative artificial intelligence models

Resumen de: GB2640912A

A method for training generative AI / machine learning (ML) models wherein the training includes iterative steps and reinforcement learning wherein a reinforcement learning reward value for one model is based on a likelihood value obtained by another model. The method includes a set of rules and iterative training steps to train the generative ML models. Each iterative training step assigns to each model a role which includes an actor model and a judge model. Each step then prompts the assigned actor model with an input to generate content that complies with a constitution. Each step also prompts the assigned judge model with content generated by the actor model and determines a likelihood of compliance that the content complies with the constitution. Each iterative training step also provides a reinforcement learning reward for training which is based on the likelihood of compliance determined by the judge model. The method may allow for the models to be trained using reinforcement learning via the reward and by switching the roles of the models. The models may be used in natural language processing tasks (e.g. reasoning, decision making etc).

PREDICTING TROPOSPHERIC DUCTING EVENTS

Resumen de: EP4648457A1

Disclosed is a method comprising collecting input data (521) comprising at least weather forecast information for an area in which one or more cells (104B, 104C, 104D) are located; providing the input data (521) to a prediction algorithm (520), wherein the prediction algorithm (520) comprises: a machine learning model (500) trained to predict tropospheric ducting events impacting the one or more cells (104B, 104C, 104D), and a cell site database (511) indicating a location and one or more configuration parameters of the one or more cells (104B, 104C, 104D); and receiving, from the prediction algorithm (520), output data (522) indicating one or more predicted tropospheric ducting events expected to impact the one or more cells (104B, 104C, 104D) based on the input data (521).

SYSTEM AND METHOD FOR MEMORY CREATION

Resumen de: EP4647971A1

The present disclosure generally relates to generating a video corresponding to a memory (e.g., an event or context) from media assets on a device. In some embodiments, the device receives user inputs requesting a video based on a natural language description of a memory. The device sends information of the natural language description to a first machine-learning (ML) model, and receives query tokens, which are used to find media items on the device that match the query tokens. The device sends information representing the found media items to another ML model that determines traits from the media items. These traits are sent to a third ML model to generate a story outline, and the video is generated by comparing the descriptions of shots in the story outline to visual embeddings of the found media assets to curate and arrange them into the video consistent with the story outline.

Input selection aware monitoring for enhanced machine learning based positioning

Nº publicación: GB2640957A 12/11/2025

Solicitante:

NOKIA TECHNOLOGIES OY [FI]
Nokia Technologies Oy

Resumen de: GB2640957A

At a first apparatus, which may be a user equipment or a gNB: receiving, from a second apparatus, which may be a gNB, a network function or an LMF, a request to initiate obtaining at least one matching factor, which indicates a degree matching between inputs used during a machine learning training phase and inputs used during a machine learning inference phase; obtaining the at least one matching factor based on the request; performing at least a machine learning model switching based on a monitoring decision request from the second apparatus, the request based at least on the obtained matching factor and at least one channel characteristic. The monitoring decision request may be further be based on at least one positioning accuracy parameter. The inputs may be transmission-reception-point measurements for a location management service or positioning determination.