Bad Election Day Forecasts Deal Blow to Data Science

By Kim S. Nash, Steven Norton and Sara Castellanos Nov 9, 2016 6:33 pm ET Big data blew it Tuesday. The most sophisticated technology in the hands of some of the smartest data scientists in the U.S. missed what grew obvious as ballots were tallied in the 2016 presidential race. Analytics experts reviewing missteps by pollsters who had predicted a strong win for Democratic candidate Hillary Clinton point to narrow data sets, faulty algorithms, and human fallibility. The findings could affect how companies conduct their own data analytics and forecasting. “You often don’t realize your model is bad until it doesn’t work,” said Thomas H. Davenport, an information technology and management professor at Babson College who studies analytics and the digital economy. Producing accurate predictions requires feeding algorithms with reliable…

Link to Full Article: Bad Election Day Forecasts Deal Blow to Data Science

Pin It on Pinterest

Share This