US Patent Issued to HITACHI on Jan. 27 for "Boosting deep reinforcement learning performance by combining off-line data and simulators" (California Inventors)

ALEXANDRIA, Va., Jan. 28 -- United States Patent no. 12,536,474, issued on Jan. 27, was assigned to HITACHI LTD. (Tokyo).

"Boosting deep reinforcement learning performance by combining off-line data and simulators" was invented by Hamed Khorasgani (San Jose, Calif.), Haiyan Wang (Fremont, Calif.), Maria Teresa Gonzalez Diaz (Mountain View, Calif.) and Chetan Gupta (San Mateo, Calif.).

According to the abstract* released by the U.S. Patent & Trademark Office: "Example implementations described herein involve an approach to address an imperfect simulator challenge using off-line data plus reward modification. The proposed solution is robust to simulator error, and therefore, it requires less maintenance in keeping the simulators updated. Ev...