Decision Science Digest: July 5, 2023

BALTIMORE, MD, July 5, 2023 –

EDITOR’S NOTE: Decision Science Digest is a periodic communique highlighting recent peer-reviewed research published by INFORMS, the largest association for the decision and data sciences, across its 17 journals. This issue highlights four press releases based on the findings of new peer-reviewed articles.

• AI-Tailor Learning: How AI Can Help Students Succeed During and After Adversity (INFORMS journal Management Science)
• Rideshare Policy Identified to Reduce Impact of Platform Critiques on Driver Revenue (INFORMS journal Management Science)
• 10 Million Fewer Cases: The Results of a Smart Testing Algorithm on COVID-19 (INFORMS journal Information Systems Research)
• Data Science Helps to Identify and Reduce Fall Risks in the Elderly Using Health Data (INFORMS journal Information Systems Research)

AI-Assisted Learning: Preparing for the Next Pandemic, How AI Can Help

The COVID-19 pandemic had noticeable impacts on education from primary through higher education. Many school districts became comfortable with virtual learning. New research in the INFORMS journal Management Science finds that artificial intelligence (AI) can help students compensate for learning loss during and after a pandemic. In the paper, “Learning Outside the Classroom During a Pandemic: Evidence from an Artificial Intelligence-Based Education App,” researchers found that immediately following the COVID-19 outbreak, students who lived in the epicenter of the outbreak used an AI education-based app less at first, but with time, they used it more, and on a more regular basis, and rebound to a curriculum path comparable to students who did not live in the outbreak’s epicenter. These findings provide understanding about how innovative education technologies can not only facilitate student learning during adversity, but also support learning recovery after adversity. Link to full article.

Rideshare Drivers are Feeling the Brunt of Platform Complaints, A New Policy Finds a Solution

Two-sided platforms match customers and service providers. Common examples include ridesharing, hotel booking, etc. New research upcoming in the INFORMS journal Management Science finds a close correlation between pricing and review systems on ridesharing platforms. More broadly, they point to the importance for these platforms to identify and address customer complaints about platform features that may have been misdirected to service providers. In the paper, “Customer Voice on Two-Sided Platforms: The Effect of Surge Pricing on Customer Complaints,” researchers find for an average trip, surge pricing makes it 1.13 times more likely for the driver to receive a complaint (either a low rating or a formal complaint). These complaints negatively impact driver’s future income. The authors also found that surge-induced complaints offset about 23% of a driver’s immediate gain from the surged fare. But after the platform adopted a policy to cap the magnitude of surge prices, the effect of surge on customer complaints reduced. The policy also led to more usage of the platform.

Smarter Testing Algorithm Could Reduce Deaths and Infection Rates if Another Pandemic Arises

New research in the INFORMS journal Information Systems Research identifies methods to proactively choose individuals to test for infection during a pandemic such as COVID-19, resulting in reduced infections and deaths. In the paper, “Smart Testing with Vaccination: A Bandit Algorithm for Active Sampling for Managing COVID-19,” researchers identify a smart testing algorithm that uses contact tracing, location-based sampling and random sampling to select specific individuals to test. Experiments based on New York City-like simulations show that smart testing can significantly reduce death rates and infection rates as compared to current practices. This algorithm resulted in a 20%-30% decline in the death rate; meanwhile, the percentage of those infected dropped by approximately 30%. In the U.S. population, these numbers are equal to approximately 10 million fewer cases and 150,000 fewer deaths. Link to full article.

New Research Helps Track Physical Data of the Elderly to Prevent Fall Injuries

New research in the INFORMS journal Information Systems Research is helping to improve the prevention of fall injuries for senior citizens based on their physical activity data. Researchers say this work can lead to significantly reduced potentially catastrophic falls by senior citizens and produce more than $33 million in economic benefits. The paper, “Motion Sensor-Based Fall Prevention for Senior Care: A Hidden Markov Model with Generative Adversarial Network (HMM-GAN) Approach,” showcases the method and explains how it works to extract temporal and sequential patterns from sensor signals for more accurate recognition of physical activities and improved prevention of falls. Using this framework, senior citizens can gain confidence and become more physically active with enhanced protection, which can in turn lead to reduced risk of psychological disorders such as depression, and improved chronic disease management. Link to full article.

About INFORMS 

INFORMS advances and promotes the science and technology of decision-making to save lives, save money and solve problems. As the largest association for the decision and data sciences, INFORMS members support organizations and governments at all levels as they work to transform data into information, and information into insights that lead to more efficient, effective, equitable and impactful results. INFORMS’ 10,000+ members comprise a diverse and robust international community of practitioners, researchers, educators and students from a variety of fields.  

###

Contact:

Ashley Smith

443-757-3578

[email protected]