A small but influential group from EY’s leadership team, including incoming Chairman and CEO Carmine Di Sibio, were on hand in a newly redesigned wavespace to recognize the winners of the EY NextWave Data Science Challenge. An extension of the program deployed in Australia last year, this global challenge resulted in 12,000 submissions from 4,500 participants from 477 universities in 15 countries.
The basic challenge: Predict human traffic patterns
The overarching goal of the project was to take a data set provided by EY partner Skyhook of citizens in the greater Atlanta area. The challenge was to take the citizens’ locations as of 3 p.m. and predict where those citizens would be located at 4 p.m. EY Global Analytics Program Director Antonio Prieto, who spearheaded this effort that will be expanded on in November, stated the intention was to connect students to a challenge that resonates with EY’s mission of building a better working world, which can be done through analytics-optimized smart cities.
Participants were allowed to enter multiple submissions as their models evolved and as they generated new “what if” scenarios. The award winners received cash prizes, EY badges and EY internships. The winners and their locations were:
First Place: Sergio Banchero is studying electronics in Australia and is a native of Brazil.
Second Place (shared): Katherine Edgley and Philipp Barthelme shared the second-place prize and are both studying applied mathematics at the University of Edinburgh.
Third Place: Chia Yew Ken of Singapore has an affinity for natural language processing and finds the parallels to AI pattern recognition interesting.
Each participant presented their basic findings and discussed the underpinning mathematical calculations and manipulations in ways that challenged this mature worker with a liberal arts background to comprehend. The incremental improvements on the algorithm scores seemed slight until put into context by Banchero, who translated his algorithm’s net improvement over the average of all submissions as ultimately capable of reducing 3,200 pounds of CO2 emissions, which would require rain forest acreage equivalent to 16 football fields to remediate naturally.