Skip to Content Skip to Content

Data Science

Coronavirus models aren’t crystal balls. So what are they good for?
Microscopic coronavirus images superimposed over digital global map

Coronavirus models aren’t crystal balls. So what are they good for?

Epidemiologists and data scientists have been gathering data, making calculations, and creating mathematical models to answer critical questions about COVID-19, but math cannot account for the unpredictability of human behavior.

Penn Medicine

Childhood exposure to trauma costs society $458 billion annually
A young child sits in a hallway burying their head in their arms on a rather dirty carpet

Childhood exposure to trauma costs society $458 billion annually

Bureaucratic hurdles block access to treatment services, so they tend to go unused. This leads to adverse outcomes that put stress on public systems like social services and law enforcement.

Michele W. Berger

What factors predict success?
A person sitting at a desk covered in papers, with a computer screen in the background. Four people are blurry, in the foreground. They are all engaged in conversation.

The findings of this latest work add to the canon of overall knowledge about what factors predict success. They also strengthen Duckworth’s original theories about grit and, at the same time, highlight other attributes key to long-term achievement.

What factors predict success?

New research from Angela Duckworth and colleagues finds that characteristics beyond intelligence influence long-term achievement.

Michele W. Berger

Removing human bias from predictive modeling
rendering of the human head from three angles with visuals of networks reaching out from the center of the brain

Removing human bias from predictive modeling

Predictive modeling is supposed to be neutral, a way to help remove personal prejudices from decision-making. But the algorithms are packed with the same biases that are built into the real-world data used to create them. 

Penn Today Staff

Can the additive tree expand machine learning in medicine?
A scan of a human body analyzed by AI tools.

nocred

Can the additive tree expand machine learning in medicine?

By combining elements of two widely used prediction models, the “additive tree” is a highly predictive model that is also easy to interpret.

Penn Today Staff

Can artificial intelligence help answer HR’s toughest questions?
A robot sits between two people at a desk, all with open laptops, the humans eye the robot suspiciously.

Can artificial intelligence help answer HR’s toughest questions?

Wharton's Peter Cappelli and Prasanna Tambe discuss the challenges companies face when they outsource their Human Resources departments to AI, allowing algorithms to remedy imperfect human decision-making for hiring, firing, scheduling, and promoting.

Penn Today Staff

Crowdsourcing 10,000 years of land use
A brown cow standing in a mountain landscape in the Italian Alps.

To predict what will happen in the future, its important to understand what happened in the past. Thats the idea behind ArchaeoGLOBE, a project that looks at land use around the world—like in the Italian Alps, seen here—during the past 10,000 years. (Photo courtesy: Lucas Stephens) 

Crowdsourcing 10,000 years of land use

More than 250 archaeologists from around the world contributed their knowledge to ArchaeoGLOBE, an effort to better understand the prevalence of agriculture, pastoralism, and hunting and gathering at different points in human history.

Michele W. Berger

Five insights into how the brain works
Person sitting at a table with blurry people in front and a screen hanging on the wall behind, which reads, "Experiential effects on brain development."

Martha J. Farah, the Annenberg Professor of Natural Sciences, is director of the Center for Neuroscience & Society at Penn. (Pre-pandemic image: Courtesy Martha Farah) 

Five insights into how the brain works

As the Center for Neuroscience & Society celebrates 10 years, founding director Martha Farah reflects on the array of research from its faculty, on subjects from brain games to aggression.

Michele W. Berger

Equifax breach and how credit agencies must change how they manage data
cartoon of person running across a full screen of 0s and 1s with an armful of numbers in the air, indicating stealing online data

After a massive data breach in 2017, the Equifax settlement with the FTC, the Consumer Protection Bureau and all 50 U.S. states calls for the firm to pay up to $700 million in damages.

Equifax breach and how credit agencies must change how they manage data

Wharton’s David Zaring analyzes the Equifax settlement, struck last week between the credit reporting firm and federal regulators over a massive data breach in 2017, and the call for stronger legislation and regulatory restraints to protect consumers.

Penn Today Staff