Analyzing destination country risk profiles in business study abroad programs: a neural network approach

Abstract

Purpose This study aims to identify how objective indicators of destination country risk differentiate business study abroad programs from those in other academic disciplines. Design/methodology/approach The authors trained a neural network model on six years of student-initiated inquiries about study abroad programs at a large US university. The model classified business versus nonbusiness study abroad programs using objective measures of destination country risk as the primary inputs. Findings The model correctly classifies business and nonbusiness study abroad programs with over 70% accuracy. Business programs were found to be 20% less likely to include destinations where the Centers for Disease Control and Prevention recommend nonroutine vaccinations and favor countries with higher Global Peace Index scores. Practical implications These results underscore the need to consider destination country risk in the design and administration of study abroad programs. An understanding of student preferences for lower risk destinations can contribute to improved planning, execution and student experiences in these programs. Social implications Better planning and management of study abroad programs based on understanding of destination country risk can lead to enhanced student safety and experiences. Originality/value This study offers a unique perspective on understanding study abroad programs by focusing on objective measures of destination country risk rather than risk perceptions. It also is, to the best of the authors’ knowledge, the first to use a neural network to classify study abroad programs as business versus nonbusiness using objective measures of country-specify risk indicators.

Research on a Highway Passenger Volume Prediction Model Based on a Multilayer Perceptron Neural Network

“…For dealing with the problem of highway passenger volume prediction, these models can effectively mine the hidden rules in the data and improve the prediction accuracy. It is of great practical significance to predict the highway passenger volume scientifically and accurately, grasp the trend, characteristics, laws, and quantity of highway passenger transport development, and make full preparations for the developmental planning of highway networks, transportation development strategy, industry management, and decision-making [3].…”