-2
$\begingroup$

After building some machine learning models in Python, R and Matlab, I found the latter's Classification Learner App to be immensely powerful. In the time it took me to build a single model in Python or R, I churned out 250 in Matlab.

Why don't auto-generated models features dominate when it comes to model creation? Also, why doesn't Matlab dominate as a result?

$\endgroup$

closed as primarily opinion-based by Matthew Gunn, Tim Jul 9 '18 at 13:32

Many good questions generate some degree of opinion based on expert experience, but answers to this question will tend to be almost entirely based on opinions, rather than facts, references, or specific expertise. If this question can be reworded to fit the rules in the help center, please edit the question.

  • 1
    $\begingroup$ Were the 250 Matlab models more accurate in prediction than the single R or Python model? $\endgroup$ – Stephan Kolassa Jul 9 '18 at 13:26
  • 1
    $\begingroup$ In addition, are you familiar with caret or scikt-lrn ? $\endgroup$ – aginensky Jul 9 '18 at 13:31
  • 2
    $\begingroup$ So what? Producing absolutely random predictions would be even faster. Ease of use, or speed are not the primary criteria for judging statistical software. $\endgroup$ – Tim Jul 9 '18 at 13:39
  • $\begingroup$ @aginensky, I hadn't come across caret. Thanks for sharing this one. But yes, I see your points that easily produced models may not offer better results than a single, well optimized model. It's something I hadn't considered. Thank you! $\endgroup$ – PizzaAndCode Jul 10 '18 at 15:04
  • $\begingroup$ I'm hearing the question as "what are the technical, production, or business weaknesses in this approach+solution in light of your limited (welcome to the human race) experience and exposure, that can cause it not to be the holy-grail (or silver-bullet or whatever). That doesn't sound so opinion based, and I think that stats folks could speak to it. They already are in the comments. Licensing fees. Alternative packages. $\endgroup$ – EngrStudent Jul 10 '18 at 15:16
2
$\begingroup$

Probably because of the high fees of MATLAB distributions, while R and Python are free.

$\endgroup$
  • $\begingroup$ I edited the question after you provided this answer, which was 'Why doesn't Matlab dominate the machine learning industry?' It was originally deemed an opinion-based question. $\endgroup$ – PizzaAndCode Jul 10 '18 at 15:08

Not the answer you're looking for? Browse other questions tagged or ask your own question.