Everyone is an expert, but a computer program may be able to choose the best


[ad_1]

UNIVERSITY PARK, Pa .– Combining human intuition with computer impartiality could improve decision-making for organizations, potentially leading to lower costs and better profits, according to a team of researchers.

In one study, researchers said that a computer program that analyzed the estimates of an agribusiness expert helped a business division of Dow AgroSciences improve the accuracy of its forecasts, resulting in a 2 to 3 increase in profits. percent and a cost reduction of 6 to 7 percent, said Saurabh Bansal, assistant professor of supply chain management at Penn State’s Smeal College of Business.

The team worked with a production expert from Dow AgroSciences management to improve forecasting in the company’s seed corn division.

According to the researchers, the production of seed corn, which farmers ultimately use as seed to produce their own crops, can be a tricky business with several factors including variations in demand and weather conditions, increasing uncertainty, according to the researchers. Researchers.

“Each year the company has to determine how many acres of land it is going to use to produce seed corn,” Bansal said. “But in this competitive industry, many varieties of seed corn are new and the company doesn’t have a lot of experience growing the new type. Therefore, she does not know what the yield would be, or how many bushels of corn they will get from her fields. However, an estimate of the yield is necessary to optimize the resources used for the cultivation of seed corn.

Companies often rely on managers as experts to provide estimates of future events and activities, as this is more cost effective than sending researchers into the field to conduct studies to gather information. However, these experts, who tend to make these predictions based on mental models drawn from years of experience, often introduce their own biases that can distort the projections.

“Everyone likes to pretend they’re experts, but deep down we know some experts are better than others,” Bansal said. “So far, there has been no objective measure of whether this expert is better than another and by how much. What we have been able to do is come up with specific measures that allow us to quantify the ‘expertise. “

The researchers, who report the results in an upcoming issue of Operations Research, available online now, developed the computer model to estimate the risk associated with return. They first collected judgments for the performance quantiles from an expert in the field. For example, the expert might estimate that there is a 50 percent chance that the business will get 55 bushels per acre.

Next, the researcher used a mathematical model to translate the quantile estimates into the mean and standard deviation of yield.

“The average provides estimates of how many bushels the business can expect on average, while the standard deviation captures the expected variability in the growth process,” said Bansal, who worked with Genaro J. Gutierrez, associate professor of information, risk and operations. management at the University of Texas at Austin; and John R. Keizer of Dow AgroSciences.

After comparing the historical data with the expert’s predictions, the program can then provide information about the bias of the expert’s own mental models, according to the researchers.

They add that using this comparison, the model quantifies expertise – or the value of expert judgments – as being equal to a specific number of data points collected in the field.

“Before that, we really didn’t know how to compare the information provided by the experts and the data,” Bansal said. “This model allows us to do just that and allows us to say that, for example, this expert is worth collecting 35 data points from field samples, which is a much more objective measure.”

He added that this is also powerful because it allows company managers to compare and select experts, determine whether to seek expert advice or collect data, as well as quantify effectiveness. of their training for experts.

Bansal said that in the future, the model could be implemented to help improve expert advice from other industries, including the biofuels industry and the semiconductor industry, which typically operate in very uncertain supply conditions.

[ad_2]

Gordon K. Morehouse