During Gordon Professor Haya Ajjan’s class on data mining, SAS data science leaders shared how analytics can be used to extract insights into coronavirus-related issues.
As part of the MGT 426 Data Mining for Managerial Decision Making course led by Haya Ajjan, the Sheldon and Christine Gordon Professor in Entrepreneurship, students are examining how data analysis techniques can be used to improve decision making.
On April 2, a team from SAS joined the class remotely through videoconferencing to discuss how analytics can be applied to gain actionable insights related to the COVID-19 outbreak.
‘Improving lives through better decisions’
Udo Sglavo, vice president of analytics research and development at SAS, reviewed how different types of data and algorithms are currently being used at SAS to support organizations and leaders in making data-driven decisions for coronavirus-related issues.
He explained SAS is using both network location and pandemic dynamics analytics to help government agencies define and deploy social distance and quarantine policies, as well as applying forecasting and optimization techniques to help hospitals and clinics optimize their use of limited resources for a high volume of critical patients. He also shared how text analytics can be implemented to quickly analyze the breadth of studies and publications about COVID-19 in order to extract insights.
“SAS as a company has been busy with trying to assist our decision makers and government institutions dealing with the coronavirus epidemic because as a company we believe everything we do should be in support of humanity…allowing people to improve their lives through better decisions,” Sglavo said.
AI for drug discovery
When it comes to identifying which chemical compounds are worth investigating as potential drugs, Jorge Silva, senior manager of machine learning at SAS, explained using a computer to go through an almost infinite number of chemicals that can be used for medicine is the more efficient option in the drug discovery process.
Silva noted how advancements in the generation of synthetic data through Generative Adversarial Networks (GANs) can help save time and cost as well as alleviate privacy concerns.
“It is a lot less controversial to use synthetic data that has the same statistical properties as real data…because of privacy concerns, especially if you are in a jurisdiction or if you need data from the jurisdiction where there are privacy restrictions,” Silva noted.
Tackling fake news
“The topic of coronavirus has become a hotbed of fake news because it’s a highly complex topic with a lot of unknowns and it elicits a strong, emotional response from people – mainly fear,” said Fijoy Vadakkumpadan, manager of computer vision at SAS. “In 2020, it’s more important than ever that we, as a society, look at this problem of fake news.”
“What the 24-hour news has really allowed those in the media to do is engage in punditry and kind of put their own spin on information,” said Lynn Letukas, director of Global Academic Programs at SAS. “It’s allowed for the misrepresentation of different images that are kind of attention grabbing or bait clicking to get you to read that particular story.”
As part an ongoing effort to tackle fake news, Vadakkumpadan demonstrated how SAS’s platform uses a collection of analytical indicators to identify and flag articles as potential fake news.
Image analytics can be used to detect the inclusion of stock photos in an article, a common practice linked to fake news, while text analytics and AI can be used to identify clickbait by finding disparities between the headline and the body of an article.
“If we can inform the user or inspire the user to take a moment and do some fact checking and talk to other people before they believe in this article or share this article with others…then we have done our job through analytics and that can be a major weapon against fake news,” Vadakkumpadan said.
Letukas encouraged the students to do their own research on COVID-19 and to continue to build their analytics skill set.
“You really have the opportunity to shape the future of analytics, whether it’s in healthcare working for a pharmaceutical company or the government, helping to solve problems like pandemics, or maybe it’s in another area like economics or business,” Letkas said.
“Hearing from data science leaders at SAS was invaluable,” Liam Lindy ’21, a computer science and economics double major, said. “Hearing how others are reacting to and learning from this crisis puts our education into context. The work that the data scientists are doing is super applicable to what we are learning in class, and allows us to remotely immerse ourselves in the crisis.”