The rise of machine learning-derived technologies means new opportunities and projects for marketers—and a massive job gap on the horizon. Now is a good time to start training for employers needing talent that can marry AI and data science with analytical chops.
First, Some Background
It’s important to note that data science and machine learning are not the same things. Data science is a massive field dealing with the intersection of data sets, computer programming, and analytics; machine learning is a subset of data science that deals with teaching computers and other machines how to learn.
Vince Lynch, the CEO of AI firm Iv.ai, recommends that marketers think about how machine learning benefits their organization before coming up with a learning strategy. “One of things that’s important is to look at the overall construct of the problem that you are facing and then think about all the data that comes into play to solve that problem,” Lynch says. For instance, if a machine learning project is based around understanding how customers interact with an ecommerce website, organizations should think about which data would be applied to the project and how it would be handled.
Companies working on artificial intelligence and machine learning projects don’t have enough talent to go around, which means workers with AI backgrounds are paid very well—between $300,000 to $500,000 if stock options are taken into account, according to the New York Times’ Cade Metz.
These high salaries have just as much to do with competition for qualified artificial intelligence experts as they do with the difficulty of the work.
A 2017 IBM report gives a good idea of the sheer numbers involved. By 2020, report authors Steven Miller and Debbie Hughes project 364,000 new jobs total created for data and analytics talent, with approximately 62,000 of those consisting of data scientists. That’s a lot of jobs.
Prepping Marketers To Learn AI
For marketers, machine learning may be trickier to learn than other data science concentrations. Potential students without a background in computer programming and mathematics will want to keep up by learning the basics of algebra, algorithms, the programming language Python, statistics and calculus. through Khan Academy or another site helps greatly with the necessary building blocks.
Lynch recommends that once marketers understand the basics, they start experimenting with different machine learning projects. “Start building things and deploying them,” he says. “Take some of the approaches that you learned about and start testing. Get data into shape, run it through models, look at the output and see how it performs.”
MOOCS, Online Courses And Self Starters
Learning the basics of machine learning doesn’t have to cost anything.
The rise of Massive Online-Only Courses—a.k.a. MOOCs—has led to a profusion of free, high-quality and in-depth online classes offering machine learning primers.
Coursera offers hundreds of machine learning and data science courses, including offerings from the University of Washington and Johns Hopkins University. One of their best-known MOOCs is the 11-week Machine Learning course, developed by Stanford University and taught by Andrew Ng, Coursera’s co-founder and the former head of Baidu AI Group/Google Brain. Coursera offers this as either a free course or a $79 version, which includes a verified certificate enrollees can share with their employers and place on their resume. Over two million students have enrolled in this particular MOOC so far.
A wide range of online machine primer courses are available free or under $100 from Udemy, Lynda, and Skillshare. Udacity, another MOOC service, offers a $1000 “Machine Learning Nanodegree” that lasts six months, with approximately ten hours of study time weekly.
Class Central, an online learning aggregator, also offers an extensive list of online machine learning courses for those seeking exactly the right fit.
There’s one big caveat about these services, however. Everyone learns differently and free online courses are not a one-size fits all solution.
MOOCs and self-paced online learning services offer two major advantages: They’re free (or low cost) and allow students to learn on their own schedule. However, students need a considerable amount of initiative, time to study, and diligence in order to succeed.
In comparison to in-person learning, there is very little hand-holding in MOOCs and responsibility for completing assignments and understanding in-class material is completely on the student.
There’s also the background knowledge students need to succeed in these courses. For marketers who are used to working with a specific marketing tech skill stack, learning the basics of machine learning will be challenging and, at times, disorienting.
Extension Learning And On-The-Job Education
While MOOC certificates look good on resumes and help open the door to interviews, many learners prefer more formal certification or certification that ties in more closely with their employers’ goals. At marketing firms, this includes everything from AI-driven data analysis to streamlining marketing costs and making sense of real-time data.
Many of the large tech incumbents, such as Facebook and Google, offer their employees free courses and resources. For instance, Facebook recently launched an internal Facebook AI Academy—a combination of in-classroom courses and on-the-job immersion designed to quickly bring engineers up to spec on various artificial intelligence-related technologies. Amazon takes this one step further and also offers machine learning and deep learning certifications for anyone working inside the AWS ecosystem.
When choosing a machine learning program, students should be prepared to shop around—check up on the bonafides of instructors, make sure their learning style aligns with what they’re looking for, confirm that the syllabus aligns with their career growth expectations, and make sure that students have used their degree or certificate for career growth afterwards.