Deloitte: 5 Trends That Will Drive Machine Learning Adoption


Companies across industries are experimenting with and using machine learning, but the actual adoption rates are lower than it might be seem. According to a 2017 SAP Digital Transformation Study, fewer than 10% of 3,100 executives from small, medium and large companies said their organizations were investing in machine learning. That will change dramatically in the coming years, according to a new Deloitte report, because researchers and vendors are making progress in five key areas that may make machine learning more practical for businesses of all sizes.

  1. Automating data science

There is a lot of debate about whether data scientists will or won't be automated out of a job. It turns out that machines are far better at doing rote tasks faster and more reliably than humans, such as data wrangling.

"The automation of data science will likely be widely adopted and speak to this issue of the shortage of data scientists, so I think in the near term this could have a lot of impact," said David Schatsky, managing director at Deloitte and one of the authors of Deloitte's new report.

Industry analysts are bullish about the prospect of automating data science tasks, since data scientists can spend an inordinate amount of time collecting data and preparing it ready for analysis. For example, Gartner estimates that 40% of a data scientist's job will be automated by 2020.

[Read the rest of this article at InformationWeek.com]

Lisa Morgan, Freelance Writer

Lisa Morgan is a freelance writer who covers big data and BI for InformationWeek. She has contributed articles, reports, and other types of content to various publications and sites ranging from SD Times to the Economist Intelligent Unit. Frequent areas of coverage include big data, mobility, enterprise software, the cloud, software development, and emerging cultural issues affecting the C-suite.

New Year's Resolutions of the Data Team

As you look back on 2017, we're sure there are improvements you would want to make. Now, as we head into 2018, what changes are you planning?

Deloitte: 5 Trends That Will Drive Machine Learning Adoption

Machine learning isn't as widely adopted as some may think, mainly because there are serious barriers to adoption. Researchers are making progress in reducing those barriers.


Re: Machine Learning
  • 12/31/2017 11:16:36 PM
NO RATINGS

Interesting! Data clean up sounds like the perfect task for machine learning. Is it common to use ML to clean data?

Re: Machine Learning
  • 12/31/2017 6:38:18 PM
NO RATINGS

"...Could machines help with that thankless task, or would we have just as much (or more) work to do after the machines finished with it ? "

 

Great question PC, And I agree data cleanup is an area where ML could help significantly.  I sometimes lose sight of the sheer size of the data blocks associated with true Big Data based projects.

Without machine assistance with these types of projects  - Analysis may very well  be impossible to carry out in a life time.

Re: Machine Learning
  • 12/26/2017 1:26:56 PM
NO RATINGS

Lisa writes: "I can tell you the big consultants are really pushing forward on a lot of automation stuff, big time, even internally.  "

Maybe I need to emphasize that I wasn't contesting that data scientists' jobs couldn't eventually be autonated, I was expressing skepticism about the pace of change, and whether 40% of these jobs would be automated by 2020.  

 

Re: Machine Learning
  • 12/25/2017 8:16:30 AM
NO RATINGS

My auto technician says that even with the diagnostic tools they have that plug into the cars there's a need or a mechanic to interpret the repair codes and determine which are leading to the correct diagnosis and which are simply misleading. Apparently even with the computerization of cars it's not quite to the stage where diagnosis of problems is a simple task.

Re: Machine Learning
  • 12/20/2017 4:15:31 PM
NO RATINGS

@Lyndon_Henry, from where I sit, I've observed that it's very easy for people to point fingers and say their company or client can automate sales, CSRs, data scientists, lawyers, researchers, etc. My question back to them is always the same:  what happens when your employer or client announces they're automating YOU out of a job?  Guess what kind of response I get: none whatsoever.

Right now, I see bullish YES camps and bearish NO camps.  What about the reasonable camp?  What's actually reasonable?  To what extent now, next year, and two years from now?  

I can tell you the big consultants are really pushing forward on a lot of automation stuff, big time, even internally.  When I asked EY Chief Innovation Officer Jeff Wong how his company was preparing employees for all this automation they're pursuing.  His response was, "We're teaching our employees how to learn how to learn."

Re: Machine Learning
  • 12/20/2017 4:02:03 PM
NO RATINGS

In her blog post, Lisa reports that

Industry analysts are bullish about the prospect of automating data science tasks, since data scientists can spend an inordinate amount of time collecting data and preparing it ready for analysis. For example, Gartner estimates that 40% of a data scientist's job will be automated by 2020.

Data scientists aren't so sure about that, and to be fair, few people, regardless of their position, have considered which parts of their job are ripe for automation.

Roughly 2 years from now, 40% of the average data scientist's job will be automated? Including data identification, collection, preparation, digitization if necessary, and processing? I have to agree with the skepticism of the data scientists on this point.  

 

Re: Machine Learning
  • 12/20/2017 12:47:12 PM
NO RATINGS

I'm just thinking out loud.  Sometimes models are used to evaluate other model.  It seems to me that it would be possibl to detect bias based on patterns.  I would see this initially as more of an alert.  It's kind of, "are you aware of this particular pattern (potential bias)?"  If effective, then clearly the concept would evolve.  Sounds like a good open source project.  I'm sure someone has thought of this...

Again, just thinking out loud.

Re: Machine Learning
  • 12/16/2017 9:41:39 AM
NO RATINGS

Bias, wherever it’s found, is difficult to objectively quantify and eliminate.

I agree with your point that that ML can make bias worse. (See Tay, the Microsoft chatbot which ‘learned’ to imitate undesirable behaviors from Twitter.)

Are you also saying that ML can held detect bias in data? Thats more difficult for me to see. Tell us more.

Re: Machine Learning
  • 12/15/2017 10:45:40 AM
NO RATINGS

I think automated bias detection tools would be interesting.  Bias is a huge issue, generally, and magnified by ML.  Things can go awry much faster, in other words.

What if there was an informative tool that ID'd bias in data and brought that to your attention before you fed that into a machine for ML purposes?  It would be kind of a safety check so, for example, Amazon doesn't break the news to me that I have a four-year-old. (A true, recent story.)

Machine Learning
  • 12/14/2017 11:48:14 PM
NO RATINGS

ML is best applied to highly repetitive tasks. Helping diagnose or repair cars - there’s millions of them and the problems are common - is a good ML problem.

Thinking about this in the context of helping data scientists, I would look at the data cleanup. Could machines help with that thankless task, or would we have just as much (or more) work to do after the machines finished with it?

INFORMATION RESOURCES
ANALYTICS IN ACTION
CARTERTOONS
VIEW ALL +
QUICK POLL
VIEW ALL +