Focus on AI, Machine Learning application development yields base for next generation of data scientists
The shortage of data scientists that was “becoming a serious constraint” according to a 2012 article in Harvard Business Review, never became a major crisis largely due to the explosive growth in data-science training during the time since.
A rush of new training programs from universities and corporations helped turn thousands of budding data scientists into the real thing according to an Aug 11, 2016 article by Tom Davenport, a senior advisor for Deloitte Analytics and co-author of the HBR piece.
There is still some debate whether data science is actually the “sexiest job of the 21st Century,” as the HBR headline phrased it. Recruiting site Glassdoor ranked Data Scientist at No. 1 in its list of the 50 Best Jobs in America, for its combination of a high median base salary, job satisfaction rate and number of available jobs.
Data scientists are likely to remain in demand, however. IDC predicts that spending on cognitive computing solutions — such as machine learning and artificial intelligence — will grow from $8 billion worldwide in 2016 to more than $47 billion in 2020. Applying data science in a business context in order to produce the kinds of insights businesses could use in both tactical and strategic decision making also helped broaden the field to include analytics experts with a strong business background, according to Bob Rogers, Chief Data Scientist for Big Data Solutions at Intel, speaking in a Forbes interview last year.
Like other organizations, Intel is responding to that growth in demand by raising the supply of training opportunities for data scientists in both academic and practical venues.
The company will sponsor the Data Strategy for Business Leaders at the University of California, Berkeley May 15-17, 2017, for example to expand the executive-education resources available to those who increasingly rely on data science.
Intel Nervana AI Academy continues to offer courses in the science and application of artificial intelligence. Intel also offers training in conjunction with four universities though the Intel Student Developer Program for AI.
The online versions of the programs over primers and tutorials on the basics on data science through courses such as AI Academy’s Deep Learning 101 webinar. Courses advance in to deeper material aimed at developers interested in expanding their skills in deep learning, machine learning, neural networks and the application of the tools including the Intel Deep Learning SDK, Intel Nervana Platform and Neon Framework.
Developers will also get the chance to get their hands on the latest Intel technology in a series of conferences and workshops though the spring and summer in locations across the U.S. and Europe.
Intel plans to expand the program during 2017 to include hundreds of academic institutions as well as a series of regional meetings, camps and workshops that will be led by experts from Intel and its partners.
It will also expand the data-science-savvy of Intel’s own employees through training and mentorship programs and a forum called the Intel Data Science Center of Excellence (DSCoE) that allows internal experts to share their skills, develop strategies and build tools applicable to the larger community.
Business-intelligence analysts can be great candidates for data-science training, but they have to be able to go beyond the pre-existing restrictions of SQL–based tools and structured data that’s already been organized, according to Intel’s Bob Rogers, who wrote a pair of blogs on career development and training of data scientists.
The most efficient approach to developing in-house data-science skills is to build a team from among existing staff, give them opportunities to expand their skills demonstrate the value of advanced analytics.
Rather than isolating them in one unit or on one set of questions, however, Rogers wrote, data-scientists should form a “SWAT Team” of people who are expert, but not overspecialized, so they can apply their skills to many different types of problems across the organization.
Their success will help evangelize the value of data science, attract more interest from BU managers and from those who want to become data scientists themselves. That, Rogers wrote, will go a long way toward relieving any shortage an individual organization has in data science, while helping employees build the skills they’ll need to lead the company into a future increasingly dependent on analysis.