United States

Amazon Patent for Data Shuffle Optimization

The USPTO has granted Amazon Technologies, Inc. a patent (US12585961B1) for a data shuffle optimization algorithm designed to reduce operators in neural network models. This innovation aims to optimize neural network models by identifying and replacing specific data shuffle regions with more efficient affine shuffle operators.

USPTO Patent Grant for ML Hyperparameter Tuning

The USPTO has granted a patent (US12585960B2) to International Business Machines Corporation for a method of dynamically tuning hyperparameters during machine learning model training. The patent describes a system that automatically adjusts hyperparameter tuning strategies based on computing resource quotas and usage limits.

USPTO Patent Granted for CDF Estimation System

The USPTO has granted a patent (US12585957B2) to the Royal Bank of Canada for a system and method to efficiently estimate a Cumulative Distribution Function (CDF) using trained neural networks. The patent covers a computer-implemented process for calculating probabilities and generating control commands based on future event estimations.

USPTO Patent Grant: Cyclical Event Period Detection System

The USPTO has granted a patent (US12586379B2) for a system designed to detect the occurrence period of cyclical events using image monitoring and advanced signal processing techniques. The patent was filed by INVENTEC (PUDONG) TECHNOLOGY CORPORATION and includes 8 claims.

USPTO Patent Grant: Mass Image Processing Apparatus and Method

The USPTO has granted patent US12586169B2 for a mass image processing apparatus and method to JEOL Ltd. The patent describes a system that uses machine learning to improve the quality of mass spectrometry images.

USPTO Patent Grant: Mitigating Bias in Machine Learning Decisioning

The USPTO has granted a patent (US12585956B2) to SAS INSTITUTE INC. for a system and method to mitigate bias in machine learning-based decisioning models. The invention focuses on detecting bias, performing counterfactual assessments, and generating explainability artifacts for user interfaces.

USPTO Patent Grant: Minimal Trust Data Sharing

The USPTO has granted a patent (US12585955B2) to Pulselight Holdings, Inc. for a computer-implemented method of protecting confidentiality during the generation of synthetic training records for machine learning using a distributed generative adversarial network (GAN). This method aims to generate synthetic data without revealing the contents of sensitive source data records.

USPTO Grants Patent for Autonomous Vehicle Operation Simulation

The USPTO has granted a patent (US12585954B2) to Wing Aviation LLC for a method of simulating autonomous vehicle operation using machine learning-generated imagery. The patent covers a computer-implemented method for improving control system performance by generating predicted and simulated aerial imagery for UAVs.

USPTO Patent Grant: Radio Signal Identification System Learning

The USPTO has granted patent US12585953B2 to Virginia Tech Intellectual Properties, Inc. for methods and systems related to machine-learned identification of radio frequency (RF) signals. The patent covers techniques for training and deploying AI models to classify RF signals based on their characteristics and environmental factors.

Samsung Patent for Optimal Neural Network Model Generation

The USPTO has granted Samsung Electronics a patent for a method to generate optimal neural network models. The patent describes a process for identifying and retaining the most effective layers of a neural network to improve efficiency and accuracy.