6 ms·

Ask HN: Any machine learning task to solve? I'll try to do it for free.

9 points by Konstantin 13y ago

It's an experience and fun for me and a good chance for you to get a working prototype. I'm a NLP engineer and my passion is sequence processing. But all other ML tasks are fine. I'll try to solve it in several hours / few days and send you the results. I am comfortable with Python (+numpy/scipy/sklearn/theano) and C (+blas/lapack), nvcc (cuda). Please send me a description (and data) to: lightcaster _at_ gmail.com You lose nothing. I give no warranties but I'll do my best. My current timezone is GMT+4.

sadkingbilly 13y ago
Have you tried Kaggle? There are some interesting projects on there.
- Konstantin 13y ago
  yes I know. My motives are a bit different. Usually kaggle competition is like a sport - fighting for hundredth of a percent in a long run. Sometimes this is not really practical. Instead I would like to grab something from real life and make it work good enough (as it usually the case for a startup). But kaggle is a cool place anyway.
  - solve 13y ago
    That's how I think too. Find the low-hanging fruit of real problems, instead of competing for a rounding error on constrained and slightly artificial problems.
ig1 13y ago
Have you seen http://www.datakind.org/ http://www.datakind.org/ ? - connecting Data Scientists to data problems faced by non-profits.
- Konstantin 13y ago
  No I haven't seen it before. Thanks for the reference!
rfergie 13y ago
I might have something for you - will check with the boss on Monday about potential privacy issues. If I don't get back to you by Tuesday then please send me a reminder email.
- Konstantin 13y ago
  Sounds good.
DanBC 13y ago
The UH Office for the Communication of Humanitarian Affairs (OCHA) has a survey about data use. (https://www.surveymonkey.com/s/ochadatasurvey https://www.surveymonkey.com/s/ochadatasurvey) Here's their website: (http://www.trust.org/?show=alertnethumanitarian http://www.trust.org/?show=alertnethumanitarian) You should be able to come up with an interesting idea.

9 points by Konstantin 13y ago

sadkingbilly 13y ago

Have you tried Kaggle? There are some interesting projects on there.

Konstantin 13y ago
yes I know. My motives are a bit different. Usually kaggle competition is like a sport - fighting for hundredth of a percent in a long run. Sometimes this is not really practical. Instead I would like to grab something from real life and make it work good enough (as it usually the case for a startup). But kaggle is a cool place anyway.
- solve 13y ago
  That's how I think too. Find the low-hanging fruit of real problems, instead of competing for a rounding error on constrained and slightly artificial problems.

ig1 13y ago

Have you seen http://www.datakind.org/ http://www.datakind.org/ ? - connecting Data Scientists to data problems faced by non-profits.

Konstantin 13y ago
No I haven't seen it before. Thanks for the reference!

rfergie 13y ago

I might have something for you - will check with the boss on Monday about potential privacy issues. If I don't get back to you by Tuesday then please send me a reminder email.

Konstantin 13y ago
Sounds good.

DanBC 13y ago

The UH Office for the Communication of Humanitarian Affairs (OCHA) has a survey about data use. (https://www.surveymonkey.com/s/ochadatasurvey https://www.surveymonkey.com/s/ochadatasurvey) Here's their website: (http://www.trust.org/?show=alertnethumanitarian http://www.trust.org/?show=alertnethumanitarian) You should be able to come up with an interesting idea.