US Patent Issued to TOYOTA RESEARCH INSTITUTE, TOYOTA JIDOSHA on Oct. 6 for "Systems and methods for adjustment-based causally robust prediction" (American, Canadian Inventors)

ALEXANDRIA, Va., Oct. 6 -- United States Patent no. 12,755,194, issued on Oct. 6, was assigned to TOYOTA RESEARCH INSTITUTE INC. (Los Altos, Calif.) and TOYOTA JIDOSHA K.K. (Aichi-Ken, Japan).

"Systems and methods for adjustment-based causally robust prediction" was invented by Guy Rosman (Newton, Mass.), Igor Gilitschenski (Toronto), Xiongyi Cui (Somerville, Mass.) and Stephen G. Mcgill (Broomall, Pa.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Prediction training systems that rely on small variation data sets instead of training the model using large passive data sets are disclosed. The smaller variation data sets are used to add loss terms that may mimic intervention. One or more models may be included...