Abstract
The assignee for this patent, patent number 12033041, is Databricks Inc. (San Fr ancisco, California,United States).Reporters obtained the following quote from the background information supplied by the inventors:“A system for big data processing comprises a system for deplo yments of applications, configurations,one or more datasets, and model(s) used in connection with analyzing the data. Models are generallydeployed in services and applications, such as web-based services, in connection with providing esti matedoutcomes, etc. A model is generated or trained based on relationships amon g different input data. Atscale, numerous models are used to provide prediction s with different aspects of a same dataset, andeach model uses numerous relatio nships among data, and the development of such relationships is veryresource in tensive. This creates a problem for training the numerous models corresponding t o a datasetin an efficient manner and tuning the models to ensure that the mode ls continue to provide effectivepredictions.”
NETL
NSTL