Diversity produces better quality for AI

Artificial Intelligence (AI) is no longer just a projection into future uses but a part of business practices. Machine learning (ML) is a tool used by businesses for predictive modeling that is used in an array of industries, from healthcare to finance to security.

The question that businesses have to address is: Are we being careful to not misuse AI by having it reinforce human biases in the training data?

To get insight into the various factors that play into that assurance, Martine Bertrand, Lead AI at Samasource in Montreal shared her thoughts. Bertrand holds a Ph.D. in physics and has applied her scientific rigor to ML and AI.

The Source of Bias

Bertrand concurs with what other experts have pointed out: “The model doesn’t choose to have a bias,” but rather she said it: “learns from the data it is exposed to.” Consequently a data set that is biased toward a certain category, class, gender, or color of skin will likely produce an inaccurate model.

We saw several examples of such biased models in Can AI Have Biases? Bertrand referred to one of the instances, that of Amazon’s Rekognition. It came under fire over a year ago when Joy Buolamnwini focused her research on its effects.

Buolamnwini found that while Rekognition did have 100% accuracy in recognizing light-skinned males and 98.7% accuracy even for darker males, the accuracy dropped to 92.9% for women with light skin and just 68.6% accuracy for darker-skinned women

Despite the demand for its removal from law enforcement agencies, the software remained in use. Bertrand finds that outrageous because of the potential danger inherent in relying on biased outcomes in that context.

Read more in  Why Diversity is Essential for Quality Data to Train AI

Tips for Aspiring Data Scientists

Kaggle graph

Data scientist ranks as the best job for 2019 in America on Glassdoor. With a median base salary of $108,000 and a job satisfaction rank of 4.3 out of 5, plus a fair number of openings predicted, that is not surprising. The question is: What does one have to do to get on track to qualify for this job?

To find out, we looked for the advice given to those who seek to get on this career track. Much comes down to the hard skills in coding and math. But that strong computation alone doesn’t cut it. Successful data scientists also need to be able to speak to business people on their own terms, which calls for the capabilities associated with soft skills and leadership.

Read more in 12 Key Tips for Learning Data Science

Getting Women to Stay On in Tech

As more and more business and manufacturing processes revolve around technology, the demand for people with the necessary skills is growing. To assure the supply of qualified people filling those positions, we have to stop thinking in terms in terms of stereotypes and clear the way for women to get on board.

Image courtesy: Pixabay

The problem is not that women aren’t trained in science, technology, engineering, and mathematics (STEM) fields. “Women have earned 57% of all bachelor's degrees and about half of all science and engineering (S&E) bachelor's degrees since the late 1990s,” according to the latest figures from the National Science Foundation. The problem is that those percentages don’t translate into the same level of representation at work.

In fact, women are still far outnumbered at engineering positions at tech companies. You can see the numbers of engineers in actual companies updated regularly on a spreadsheet in Tracy Chou's Women in Tech list.  Though the numbers vary, the average representation for women engineers at the companies listed appears to be near 20% to 25%.

The gap between the sexes grows higher up the hierarchy. The Gender Divide in Tech-Intensive Industries put out in 2014 demonstrated that women with MBAs with tech qualifications were still far less likely to work in the industry than their male counterparts. Perhaps part of the reason is that women ae far more likely to be placed in entry level jobs, at the rate of 55% in contrast to the 39% for men.  Women MBAs also were more likely to leave the tech industry than their male counterparts at the rate of 53% to 31%.

Read more in 

Retaining Women in Tech Takes More Than Training

To certify or not to the certify: the building blocks of a blockchain career

It’s clear that tech pros in a variety of industries are examining the implications of blockchain. After all, the technology can be leveraged not only for cryptocurrency (i.e., Bitcoin), but everything from “smart” contracts to secure, distributed ledgers. Can developers prove that they have the skills to work with this technology?
Read more in Can You Obtain Certifications for a Blockchain Career?

Data science training to fill the skills gap

 Datanami article looked at various estimates about how many more data scientists from firms like Gartner and McKinsey to arrive at a range of 100,000-190,000 shortfall by 2020 and 2017, respectively. That's about 60% more demand than available supply in the US. The upside is that salaries for those jobs go up, but the downside, of course, is that the businesses cannot progress on the data front as much as they would like to without the skilled people in-house.

The obvious solution to that problem is to get more people qualified for these jobs. If they didn't learn the skills in college programs, they can now get trained in specialty programs like the NYC Data Science Academy (NYCDSA). Vivian Zhang, the CTO of the school, set it up specifically to meet industry needs and get people the data science training needed to fill those open job slots.

Read more in 

Mind The Gap: Minting New Data Science and Analytics Professionals