KDnuggets Home » News » 2019 » Mar » Opinions » Where Analytics, Data Science, Machine Learning Were Applied: Trends and Analysis ( 19:n11 )

Where Analytics, Data Science, Machine Learning Were Applied: Trends and Analysis


CRM/Consumer analytics, health care, banking, finance, and science were the top sectors in 2018. The greatest increases were in mobile apps, investment, security, entertainment, and social policy, while fraud detection, retail, advertising, direct marketing, and social media saw the greatest declines.



In the latest KDnuggets poll, readers were asked:

What are the Industries / Fields where you applied Analytics, Data Science, Machine Learning in 2018?

Notably, CRM/Consumer Analytics remains number 1, which it has been for the last 5 years. Health care made a jump to number 2 from number 4 last year. Conversely, finance dropped to number 4 from number 2; however, these are small percentage differences in the share, and so are not statistically significant.

The top 5 industries were:
  • CRM/Consumer analytics, 18.6%
  • Health care, 17.2%
  • Banking, 17.0%
  • Finance, 16.1%
  • Science, 13.6%

Figure
Fig. 1: Where Analytics, Data Science, Machine Learning were applied in 2017, 2018 - KDnuggets Polls


When compared with the results of last year's survey, the most significant increases, computed as (pct2018 - pct2017) / pct2017, were as follows:
  • Mobile apps, 4.4% in 2018, up 224.7% from 2017
  • Investment / Stocks, 6.7% in 2018, up 98.2% from 2017
  • Security / Anti-terrorism, 3.0% in 2018, up 90.4% from 2017
  • Entertainment / Music / TV / Movies, 4.8% in 2018, up 79.4% from 2017
  • Social Policy / Survey analysis, 2.5% in 2018, up 61.1% from 2017

The most significant decreases over the same period were:
  • Fraud Detection, 10.6% in 2018, down 18.7% from 2017
  • Retail, 9.4% in 2018, down 15.9% from 2017
  • Advertising, 9.0% in 2018, down 14.9% from 2017
  • Direct Marketing / Fundraising, 4.4% in 2018, down 11.5% from 2017
  • Social Media / Social Networks, 6.0% in 2018, down 11.1% from 2017

Once again, the average number of industries/areas per voter remained quite consistent. There was a small bump this year to 3.0 per participant, as compared with 2.7 in both last year's poll (2017) and 2016, and only a slight change from 2.65 in 2015. This may bolster Gregory Piatetsky's comment from last year's poll analysis: "Perhaps there is a limit to effective application of Data Science across industries?"

This year's poll had 435 participants, which was down slightly from 446 last year. There were 552 participants in the 2016 poll, and 350 in 2015.

7 sectors have shown an increase of 10% or more for 2 years in a row:
  • Health care
  • E-commerce
  • Education
  • Manufacturing
  • Automotive / Self-Driving Cars
  • HR / workforce analytics
  • Junk email / Anti-spam

while only 2 areas have seen a decline in of 10% or greater in the past 2 consecutive years:
  • Advertising
  • Social Media / Social Networks

The next page shows full results across the last 3 years.

Sign Up

By subscribing you accept KDnuggets Privacy Policy