Alerta

Response Prediction for Electronic Communications

Resumen de: US20260017544A1

Systems, methods, and apparatuses are described herein for performing sentiment analysis on electronic communications relating to one or more image-based communications methods, such as emoji. Message data may be received. The message data may correspond to a message that is intended to be sent but has not yet been sent to an application. Using a first machine learning model, one or more subsets of the plurality of emoji may be determined. The one or more subsets of the plurality of emoji may comprise one or more different types and quantities of emoji, and may each correspond to the same or a different sentiment. Using a second machine learning model, one or more emojis may be selected from the one or more subsets. The one or more emojis selected may correspond to responses to the message.

AUTOMATED FEATURE SELECTION FOR SPLIT NEURAL NETWORKS

Resumen de: US20260017517A1

A computer-implemented method and apparatus for feature selection using a distributed machine learning (ML) model in a network comprising a plurality of local computing devices and a central computing device is provided. The method includes training, at each local computing device, the ML model during one or more initial training rounds using a group of input features representing a input features layer of the ML model. The method further includes generating, at each local computing device, based on the one or more initial training rounds, feature group values. The method further includes transmitting, from each local computing device, to the central computing device, the generated feature group values. The method further includes receiving, at each local computing device, from the central computing device, central computing device gradients. The method further includes computing, at each local computing device, local computing device gradients, using the received central computing device gradients. The method further includes generating, at each local computing device, a gradient trajectory for each input feature in the group of input features based on the computed local computing device gradients. The method further includes identifying, at each local computing device, based on the generated gradient trajectory, whether each input feature in the group of input features is non-contributing. The method further includes removing, at each local computing device, from the group

COMMUNICATION METHOD AND APPARATUS

Resumen de: US20260019345A1

A communication method and apparatus. A first device sends capability information to a second device, so that the second device can send control information to the first device based on the capability information. The control information is usable to indicate that a first machine learning (ML) model corresponding to a first management function (MnF) on the first device is allowed to be trained. The first device trains the first ML model based on the control information.

MACHINE LEARNING DEVICE, MACHINE LEARNING METHOD, AND COMPUTER-READABLE MEDIUM STORING MACHINE LEARNING PROGRAM

Resumen de: US20260017931A1

A machine learning device includes an image set acquiring unit to acquire an image set including images, and an image set selecting unit to select an image set similar to the acquired image set from a plurality of image sets different from the acquired image set. In addition, the machine learning device includes a performance comparison unit, and a preprocessing acquisition unit to select a learning model from a plurality of machine-learned learning models based on a performance comparison result by the performance comparison unit and to acquire preprocessing performed on an image set used for machine learning of the learning model selected. Furthermore, the machine learning device includes a model learning unit to perform the acquired preprocessing on the acquired image set and to cause a learning model that has not yet trained to perform machine learning using the preprocessed image set.

APPLICATION OF ARTIFICIAL INTELLIGENCE MACHINE LEARNING (AIML) MODELS ASSOCIATED WITH SAME FUNCTIONALITY

Resumen de: WO2026015208A1

Disclosed are techniques for wireless communication. In an aspect, a processing device may receive, from a server device, a request for an output based on application of a plurality of artificial intelligence machine learning (AIML) models associated with a same functionality. The processing device may apply the plurality of AIML models to obtain a plurality of respective candidate outputs, the plurality of candidate outputs being associated with the functionality. The processing device may transmit the output to the server device in response to the request, the output indicating at least one of the plurality of candidate outputs.

Method for improving accuracy of machine learning models

Resumen de: GB2642421A

Method for training a neuro-symbolic machine learning model, comprising: for each image depicting at least two objects of a training dataset: inputting the image into a neural module 102, 200, 202 to obtain bounding boxes and features therein (digit); inputting each bounding box and object feature into a symbolic module (106, Fig.1; rest of Fig.2) to obtain a plurality of possible labels i.e. partial labels 212 and possible relationships 210 as a new partially-labelled training dataset; and training the neuro-symbolic model (neural module and the symbolic module) by calculating a loss from a ground truth label for the image. The symbolic module may use a set of logical rules to constrain the labels and explanations (R1-R5, Fig.7). The trained neuro-symbolic model may generate a scene graph, perform action recognition, perform visual question answering (Fig.4) or control an autonomous or semi-autonomous electronic device. The electronic device may be a moveable robot or a wearable augmented reality device.

METHOD FOR PROVIDING INTELLIGENT RESPONSE AGENT BASED ON ADVANCED INFERENCE AND ESTIMATION FUNCTION, AND SYSTEM THEREFOR

Resumen de: EP4679287A1

A method and system for providing an intelligent response agent based on a sophisticated reasoning and speculation function according to an embodiment of the present disclosure can generate and provide response data for queries related to specialized documents using a deep-learning neural network that implements a stepwise process for a sophisticated reasoning and speculation function.

MULTI-TASKING MODEL TRAINING METHOD AND MULTI-TASKING PERFORMING METHOD USING MACHINE LEARNING MODEL TRAINED ON BASIS THEREOF

Resumen de: EP4679330A1

A multi-tasking model training method and a multi-tasking performing method using a machine learning model trained on the basis thereof, according to an embodiment of the present invention, may mutually transfer and learn knowledge data of a latent space for each task through geometric alignment in one integrated latent space in order to process a multi-task for output according to a plurality of domains.

TRIGGER-BASED DATA INGESTION FOR MACHINE LEARNING USING EDGE DEVICE

Resumen de: WO2026010723A1

An edge device comprising processing circuitry and memory stores a representation of a trigger condition. The edge device accesses streaming sensor data. The edge device determines, based on the streaming sensor data and using the processing circuitry, that the trigger condition is met. The edge device transmits the streaming sensor data to a computing device in response to determining that the trigger condition is met.

ADAPTABLE, SCALABLE, AND AUTONOMOUS PROTECTION VERIFICATION AND DECISION SUPPORT

Resumen de: WO2026010799A1

A method includes obtaining (302, 802) information associated with assets and/or personnel to be protected and executing (306-322, 804) a set of weighting functions and a set of algorithms for protecting the assets and/or personnel. The weighting functions and algorithms are arranged in multiple levels of a hierarchy. Each level of the hierarchy includes one or more of the weighting functions and one or more of the algorithms. The one or more weighting functions and the one or more algorithms in at least one level of the hierarchy are applied across a timeline. The method also includes applying (330, 818) an artificial intelligence/machine learning (AI/ML) algorithm (608) across the timeline to update results due to one or more changes during one or more operations involving the assets and/or personnel.

INPUT FOR MACHINE LEARNING PREDICTION MODULE IN WIRELESS COMMUNICATION SYSTEM

Resumen de: WO2026010874A1

A UE (102) receives (306), from a network entity (104), a configuration configuring a first set of RSs associated with an ML prediction module. The UE receives (310) a second set of RSs different from the first set of RSs. The UE transmits (316) a prediction report based on a measurement of the second set of RSs being used in an input to the ML prediction module. A UE (102) receives (806), from a network entity (104), a configuration configuring a plurality of RSs associated with a plurality of ML prediction modules. The UE receives (808) an indication indicating an RS associated with an ML prediction module. The ML prediction module is being executed at the UE. The UE transmits (816) a prediction report output from the ML prediction module based on a measurement of the RS being used as an input to the ML prediction module.

SYSTEM AND METHOD FOR LEARNED EMITTER IDENTIFICATION AND TRACKING

Resumen de: WO2026010990A1

A system and method are described for emitter identification and tracking in an electronic warfare (EW) environment. The system includes an antenna array configured to receive signals from radio frequency (RF) emitters during a dwell. Processing circuitry converts the received signals into digital signals. Pulses are detected and characteristics of the pulses determined to form pulse descriptor words (PDWs). The PDWs obtained during the dwell are deinterleaved using unsupervised machine learning to form clusters. The clusters are categorized using one or more supervised machine learning algorithms to determine whether the PDWs correspond to known or unknown emitters and the results tracked as in or out of library emitters. After merging the in or out of library emitters, an emitter report is generated and used to update a library of emitter profiles used by the supervised machine learning algorithms as well as determine countermeasures to generate.

SYSTEMS AND METHODS FOR REDUCING CARBON FOOTPRINT BY TRACKING AND VERIFYING CARBON INTENSITY IN SUPPLY CHAIN OPERATIONS

Resumen de: AU2025271453A1

Examples of the present disclosure describe systems/methods of reducing carbon footprint by generating and tracking a carbon intensity (CI) score assigned to a particular product as the product traverses through a processing plant and discrete steps in a supply chain. In some examples, intermediate CI scores may be assigned to the product as it completes each step in its life cycle. The intermediate CI scores may be aggregated to produce a final CI score. Each intermediate CI score is recorded on a blockchain, such that the CI score is independently verifiable and auditable. In other example aspects, a machine-learning model may be applied to the input data received from each supply chain stakeholder and CI scores, wherein the machine-learning model generates intelligent suggestions to stakeholders for how to tweak their processes to lower CI scores. In other examples, a CI score may be used to derive a value for a CI token. ov o v

Trigger-Based Data Ingestion for Machine Learning Using Edge Device

Resumen de: US20260012762A1

An edge device comprising processing circuitry and memory stores a representation of a trigger condition. The edge device accesses streaming sensor data. The edge device determines, based on the streaming sensor data and using the processing circuitry, that the trigger condition is met. The edge device transmits the streaming sensor data to a computing device in response to determining that the trigger condition is met.

SYSTEM AND METHOD FOR LEARNED EMITTER IDENTIFICATION AND TRACKING

Resumen de: US20260009880A1

A system and method are described for emitter identification and tracking in an electronic warfare (EW) environment. The system includes an antenna array configured to receive signals from radio frequency (RF) emitters during a dwell. Processing circuitry converts the received signals into digital signals. Pulses are detected and characteristics of the pulses determined to form pulse descriptor words (PDWs). The PDWs obtained during the dwell are deinterleaved using unsupervised machine learning to form clusters. The clusters are categorized using one or more supervised machine learning algorithms to determine whether the PDWs correspond to known or unknown emitters and the results tracked as in or out of library emitters. After merging the in or out of library emitters, an emitter report is generated and used to update a library of emitter profiles used by the supervised machine learning algorithms as well as determine countermeasures to generate.

SYSTEMS AND METHODS FOR FIXED ROUTE BUS SPEEDS FOR A FLEET OF BUSES

Resumen de: US20260011243A1

A system and method for predicting fixed route travel time (e.g., bus speeds along bus routes) is provided. The system and method include a first machine learning model trained to predict speed along the fixed route without turning and dwell times. The speed from the first machine learning model, along with dwell time and turn time can be used with a second machine learning model to determine the overall route time.

ARTIFICIAL INTELLIGENCE-ASSISTED BUILDING AND EXECUTION OF A FEDERATED DATA LAYER FOR ENTERPRISE ENGINEERING

Resumen de: US20260010543A1

Artificial Intelligence-assisted building/execution of federated data layer for enterprise engineering: A system trains at least one machine-learning model to identify information about industrial assets from training data, then map the information to a federated data model. The system retrieves information about data from an application in an industrial asset. The at least one machine-learning model identifies types of the data, relationships between the data, and patterns of the data, from the information and based on data types, data relationships, and data patterns in the federated data model. The at least one machine-learning model maps the types of the data, the relationships between the data, and the patterns of the data to the federated data model. The system identifies knowledge about the types of the data, the relationships between the data, and/or the patterns of the data in the federated data model, in response to a query about data.

METHOD FOR ASSESSING THE SEISMIC RISK ON EXISTING BUILDINGS

Resumen de: US20260009912A1

Method for assessing the seismic risk on existing buildings, comprising the following steps: a) identifying a set (N) of existing buildings to assess;b) acquiring for all existing buildings belonging to said set (N) qualitative data relating to the formal and construction features of said buildings;c) processing said qualitative data with a rapid analysis method based on qualitative criteria to assess the seismic vulnerability, and the related basic seismic risk, of all existing buildings belonging to the set (N);d) selecting in an organized manner a subset(S) comprising 25% to 33% of buildings of the set (N);e) acquiring for all the buildings of the subset(S) a plurality of analytical parameters;f) processing said plurality of analytical parameters with a scientific analysis method based on quantitative criteria to assess the vulnerability and the basic seismic risk of all the buildings of the subset(S);g) selecting in an organized manner a learning sample (A) comprising 70% to 80% of the buildings of the subset(S), and deriving by subtraction a verification sample (V) comprising 20% to 30% of buildings of the subset(S);h) using an AI-based machine learning model entering into an algorithm, for each building included in said learning sample (A), at least a part of said plurality of analytical parameters and the corresponding seismic vulnerability and basic seismic risk results already obtained with the scientific analysis method referred to in step f), to generate a statisti

SLICE-BASED METHODS FOR OPTIMIZING VALIDATION AND SUBSEQUENT RETRAINING PROCEDURES OF MACHINE LEARNING MODELS

Resumen de: US20260010835A1

Methods for a machine-learning network that provide efficient, scalable, and granular analyses during validation of a machine learning model are disclosed. Providing quantitative analysis information to machine learning experts when they are deciding how to proceed with further optimizing their given machine learning model allows for more directed procedures during edge case detection. Following the execution of a principal machine learning model using a validation dataset, a shallow learning model may be trained to provide simulations about how the principal model may be improved, or not, given different re-training scenarios. By using slice-based schemes, the validation dataset is divided into certain problematic data slices, and then, during inference of the shallow learning model, additional quantitative information about the effect on other data slices given a subsequent re-training of the principal model using a certain problematic data slice allows the ML expert to make more informed decisions.

DETECTING MALICIOUS EMBEDDINGS IN DOCUMENTS DESTINED FOR NETWORKED SYSTEMS

Resumen de: US20260010624A1

The technology relates to cybersecurity attacks and cloud-based security, and specifically to detecting malicious embeddings in document destined for a networked system. Such embeddings can be delivered in the form of malicious macros and/or malicious OLE objects stored within document files. The technology intercepts a document that is compatible with an MS Office file format, finds embedded code, engineers at least five features that characterize the embedded code. The technology inputs the engineered features to a trained machine learning model and determines, as a threat level, a likelihood of malicious embedding from at least the engineered features of the embedded code. Based on the threat level, the technology can block the document with a malicious threat level, accept the document with a non-malicious threat level, and or isolate the document with a suspicious threat level.

SMART RING SYSTEM FOR MONITORING UVB EXPOSURE LEVELS AND USING MACHINE LEARNING TECHNIQUE TO PREDUCT HIGH RISK DRIVING BEHAVIOR

Resumen de: US20260010807A1

A method for predicting risk exposure can include receiving data from a sensor. The method for predicting risk exposure also can include analyzing the data via a machine learning (ML) model. The analyzing can include determining that the data represents a light exposure pattern correlated with a risk pattern. The ML model can be trained with training data indicative of the light exposure pattern and indicative of the risk pattern to identify a correlation between the light exposure pattern and the risk pattern. The method for predicting risk exposure further can include predicting a risk exposure for a user based on the analyzing the data. The method for predicting risk exposure further can include providing a notice indicating the risk exposure, as predicted. Other embodiments are disclosed herein.

DEVICE, SYSTEM, AND METHOD TO GENERATE RECOMMENDATIONS FOR FIXING MALFUNCTIONS IN A VEHICLE

Resumen de: WO2026009195A1

The present invention discloses a device (100), a system (300), and a method (200) for generating recommendations for predicting malfunctions in a vehicle (500). The invention includes a device (100) for generating recommendations after detecting vehicle (500) malfunctions (500). The device (100) comprises a control unit (101) with a transceiver (103) receiving sensed parameters from vehicle sensors (102) and user inputs. The processors (101-1) within the control unit (101) analyze the combined data to create the datasets to identify faults and predict potential malfunctions. The control unit (101) can further employ a fault identification model (104) like machine learning or artificial intelligence models to aid the analysis. Based on the analysis, the device (100) generates recommendations for users to address potential vehicle (500) malfunctions.

METHODS, SYSTEMS AND COMPUTER PROGRAMS USING MACHINE LEARNING TO OPTIMIZE PREDICTION OF AN OCCURRENCE OF A RECURRING MEDICAL SYMPTOM OR BODY BEHAVIOR

Resumen de: WO2026008166A1

A computer-implemented method for optimizing a prediction of an occurrence of a recurring medical symptom or body behavior of a human being using machine learning, the method comprising obtaining (110) input data comprising features related to the occurrence of the recurring medical symptom or body behavior, dividing (120) the features included in the input data into two or more groups of features, encoding (130), using a trained encoder machine learning model, the features of the two or more groups of features into two or more embeddings, with each embedding representing a group of features, inputting (150) the two or more embeddings into a prediction machine learning model being trained to predict the occurrence of a recurring medical symptom or body behavior based on the two or more embeddings, and providing (160) a prediction of the recurring medical symptom or body behavior based on an output of the prediction machine learning model.

METHOD AND SYSTEM FOR EVALUATING LIFE OF BATTERY THROUGH AI-BASED ELECTRIC VEHICLE OPERATION DATA ANALYSIS

Resumen de: EP4674678A1

According to one embodiment of the present invention, there is provided a battery life evaluation system including: a data collection unit configured to obtain information about route information, charge/discharge patterns, driving patterns, and cell states from an electric vehicle; and a life evaluation unit configured to predict the life of a battery installed in the electric vehicle by performing machine learning based on the obtained information.

METHOD FOR ASSESSING THE SEISMIC RISK ON EXISTING BUILDINGS

Nº publicación: EP4675519A1 07/01/2026

Solicitante:

M T RICCI S R L [IT]
M.T. Ricci S.r.l

Resumen de: EP4675519A1

The method for assessing the seismic risk for existing buildings introduces an integrated and codified flow of work steps through a mixed qualitative-quantitative assessment with the combination of the expeditious method with the scientific method, performed on a strategic subset of the population, an operational and targeted use of artificial intelligence with the use of supervised machine learning models, validated on an empirical basis, with APS (accuracy, precision, sensitivity) metrics calculated on a verification sample, with a significant reduction in time with the application of the validated model to a remaining subset S' (67-75% of the building heritage) without the need for a complete engineering assessment, offering a modular structure which may be adapted to different territorial and typological contexts, with the selection of the most relevant variables (P1/P'1, P4/P'4, P6/P'6, P7/P'7) performed through the ANOVA statistical analysis and the Chi-Square analysis, also thanks to a systemic approach with the definition of a clear, replicable operational flow, scalable and provided with objective validation criteria.