Positive reinforcements help algorithm forecast underground natural reserves

Researchers have designed a reinforcement-based algorithm that automates the process of predicting the properties of the underground environment, facilitating the accurate forecasting of oil and gas reserves.