Undergraduates and graduates alike are learning “magic” from Professor of Information Systems Manoochehr Ghiassi. While the new dynamic artificial neural network model (DAN2) Ghiassi has developed is not exactly a crystal ball, its innovative use of artificial intelligence to accurately predict upcoming operational needs based on current data helps businesses forecast the future and use their resources more efficiently.
“Businesses use data, model it, and analyze it in order to make better decisions for their future operations. My students are learning how to do that, right now while they’re students,” Ghiassi says.
For example, he has adapted his computer algorithm to accurately forecast monthly, weekly, daily, and hourly urban water demand for the San Jose Water Co. with an accuracy ranging from 95 to 98 percent.
In his lab, seven students assist him in running experiments to hone computer models for different applications. “I teach them about DAN2 models and then they get to a point where they can help run the experiments. They run an experiment. They bring the results to me. We sit down and discuss it and decide how to do the next one, ” he notes. With Ghiassi’s guidance, students of today are learning to apply the information systems tools of tomorrow.