Subscribe to DSC Newsletter

What are the most difficult things to predict, among these 19 challenges?

Please share your opinion and let us know how to make better predictions for the following problems. Also how do you define better?

  1. Earthquake time and intensity forecasts
  2. Solar flares time and intensity forecasts
  3. Stock market prices for individual stocks or indexes, buy/sell signals
  4. Efficiency of new pharma drugs (in terms of sales, or in terms of health improvement)
  5. Weather forecasts in UK, the NorthWest or other challenging locales
  6. Bankruptcy prediction
  7. Econometric: how to predict the success of a new immigration policy, or a new tax regulation, or a new law to stop drug abuse or crime
  8. Asteroid collisions with earth
  9. Early detection of a pandemic
  10. Chance for a borrower to repay her loan
  11. Terrorist attacks
  12. Chance of a 5 or 10 or 20-year survival for a particular person
  13. Spam detection
  14. Bogus reviews detection
  15. Chance for an ad to result in a sale
  16. Risk of car accident within the next 12 months for a particular individual
  17. Chances of succeeding at University based on high school metrics
  18. Chances of succeeding in a job based on job interview metrics
  19. Chances of succeeding for a start-up, if you invested $10MM in the start-up in question

Views: 5283


You need to be a member of AnalyticBridge to add comments!

Join AnalyticBridge

Comment by Shaizan on August 31, 2012 at 5:39am

It would be interesting to predict #16 Risk of car accident within the next 12 months for a particular individual

Comment by Benhenni on August 29, 2012 at 7:06pm

Regarding #11, I wasn't able to find the references, but I read about models doing correct predictions in few cases, giving probabilities of occurring events in a given area considering what happened in the neighboring. I know such models are now used somewhere in the United States by police officers to to "optimize" their patrol around predicted hot points.

Comment by Alfred on August 23, 2012 at 10:30am

I think economics and medical problems are the hardest, so I would say 7 then 6 12 and 19, since they are affected by non-measurable parameters.

I would add a last one thing to reach a round figure :

20. Long Term Demographic projections...?

Comment by Matthew A. Riebel on August 20, 2012 at 5:25pm

I hope it's 18 because I was continually rejected on interviews for a long time based on my social struggles even though I knew I would perform the job very well if given the opportunity.

I also think 11 is hard to predict.  Consider how unprecedented September 11th was.

Comment by Nissim Matatov on August 19, 2012 at 5:10am

I think it depends on organization . Each time the challenge lie in different part of analytics solution : data or algorithm issue. I mean , for example , paragraph 10 can be the easiest and hardest in two different organizations.

You can talk about aggregated measure based on existing practices . May be it's worth to run an appropriate survey.

On Data Science Central

© 2021   TechTarget, Inc.   Powered by

Badges  |  Report an Issue  |  Privacy Policy  |  Terms of Service