Valuable Outputs from 'Women in Data Science' Event at Stanford

Last week's reportage on the Microsoft Learning Born to Learn blog included a simply stellar piece from Carolyn Lesser, Engineering Director for the Learning Experiences (LeX) Platforms team at Microsoft.

In a Feb. 15 post titled "Insights from Stanford's Women in Data Science Conference," Lesser gives an overview of what went down in Palo Alto (as well as at more than 70 satellite events around the globe). It sounds like a veritable treasure trove of information, strategy, and explanations of work in progress was tendered to a large world-wide audience.

I'll cover some fascinating highlights here, but I urge everybody to consult the afore-linked blog post and visit the WIDS website. Anyone with an interest in data science will want to take full advantage of all the information to be mined there.

As Ms. Lesser observes, Data Science is a huge and growing field nowadays, with event attendees coming from backgrounds in engineering, product planning and sales, just at her virtual conference table. Others who participated represent software development, mathematics, and various technical fields as well.

As Lesser sees it, the impetus toward data science is simple: "Data is good business sense, and makes us smarter in our decisions." Microsoft itself has invested heavily in data science over the past few years and, in fact, expanded its Microsoft Professional Program to include Data Science in 2016.

To summarize her take on things, Ms. Lesser sees cloud computing as lying at the center of the upswing in data sciences. This applies particularly in the areas of machine learning and artificial intelligence, because of the Cloud's ability to deliver more and bigger data sets, as well as huge amounts of computing capability, to would-be consumers of such things anywhere, anytime.

She explains how this is affecting artificial intelligence, which used to rely on painstakingly analyzed and carefully crafted sets of rules to enable systems to interact with the real world. In an age of machine learning and insane amounts of data, however, along with automated learning algorithms, systems and machines can now teach themselves, and build their own sets of rules, algorithms, and guidelines.

Another enthralling topic in her discussion came from the work of Miriah Meyer, an assistant professor at the University of Utah, who gave a talk on the critical role that visualization plays in research, both during the process of scientific discovery, and then later in the communication of results to others.

Effective visualization depends on numerous factors, including the right kinds of visual channels, so that number data may be enriched and enhanced through careful use of color, density, volume, area, angle, slope, length, and position. Determining how to understand such data is essential to defining actionable tasks.

The overall process of creating and delivering effective visualizations is something Meyer calls "data counseling," which means equating a task with an action, an object, and a descriptor. The action is what you want to do, the object is what you want to act upon, and the descriptor is a value to measure to gauge the results.

Ms. Lesser also elaborated on how Microsoft works with business stakeholders to understand reporting requirements and to create "insight hacks" to measure and evaluate a business. Her group is applying data sciences to better understand (and where possible, to replicate) learner behavior and outcomes, particularly successful ones.

The same techniques also help her group to keep Azure costs under control, to assess customer sentiment and reactions, and to improve the overall experience delivered by Microsoft Learning.

Finally, Ms. Lesser presented a rubric of essentials that a panel of distinguished women leaders offered to help interested professionals learn about and develop data science skills:

? Analytical thinking
? Statistics/math
? Coding (scripting, etc.)
? Product or business sense
? Communication (tell the story)

It's important to note that the real value of data science comes from the intermingling of hard-boiled number crunching and analytical insight, along with squishier abilities to draw insight from a sense of the market and the business context in which one operates (as well as the ability to share the insights gained thereby with others).

This speaks compellingly to the magic combination of hard and soft skills that makes for superb and effective technologists.

There's a lot to gain from learning about data science. I hope you'll consult the various resources linked here and in my other posts on Big Data and data science to do just that!

Here's a small sampling:

? Who's Working in Big Data Nowadays? (June 12, 2015)
? Continuing Opportunities in Big Data for IT Pros (Nov. 27, 2015)
? CompTIA Says Big Data Exceeding Expectations (Dec. 18, 2015)
? LinkedIn Dives into Data, Finds Leading Tech Skills for 2016 ( Jan. 22, 2016)