The Entrepreneur Forum | Financial Freedom | Starting a Business | Motivation | Money | Success

Welcome to the only entrepreneur forum dedicated to building life-changing wealth.

Build a Fastlane business. Earn real financial freedom. Join free.

Join over 80,000 entrepreneurs who have rejected the paradigm of mediocrity and said "NO!" to underpaid jobs, ascetic frugality, and suffocating savings rituals— learn how to build a Fastlane business that pays both freedom and lifestyle affluence.

Free registration at the forum removes this block.

Data Science / Machine Learning Contests

ZeroToOne

Contributor
Read Fastlane!
Speedway Pass
User Power
Value/Post Ratio
157%
May 6, 2017
44
69
KS
To the data scientists and machine learning/AI enthusiasts out there, have you had any experience with Quantopian or QuantConnect or know someone who has? I'm specifically interested in the format of the contests on those sites; where people submit trading algorithms that are back tested and then scored on out-of-sample data. That competition model (testing ML algorithms against each other) might work well for the business model I'm currently working on.

I'm specifically looking for:
  • Input/guidance on setting up / running the competitions
  • Are data scientists more likely to use Python, R, JavaScript, etc?
  • Are there any other non-trading competitions with a similar model?
  • Suggestions on the type of person should I be looking at hiring to help. I have a professional programming background, and have even played with machine learning (Keras / TensorFlow), but I do not have any sort of formal data science experience.
Any input would be helpful. Thanks!
 
Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

ChrisV

Legendary Contributor
EPIC CONTRIBUTOR
Read Fastlane!
Read Unscripted!
Speedway Pass
User Power
Value/Post Ratio
225%
May 10, 2015
3,141
7,055
Islands of Calleja
Quantopian or QuantConnect
Hmm.. I'm not familiar with Quantopian or QuantConnect specifically, but Kaggle is the go-to place for Machine Learning competitions.


Data scientists are most likely to use R or python. Which one you choose depends on what you want to do: R is better for Academia, python is better for general-purpose. If you look at the numbers I believe python is now the leading language for Data Science. Also, the field is outgrowing both R and python and I believe it's going to move toward Swift. If you learn python you'll be in a much better position to learn Swift since they're very similar.

Unless you want to work in academia, I suggest python. If you want to work in Academia, learn R. Most academics use R and there's actually a stigma against python in Academia.

I personally use R, but I've been picking up some python as well.

Javascript I don't recommend.

What do you mean by hiring? To help you learn it?
 

ZeroToOne

Contributor
Read Fastlane!
Speedway Pass
User Power
Value/Post Ratio
157%
May 6, 2017
44
69
KS
I think in my use case, academia or general-purpose is fine. The result of the programs are all I'm really interested in, not necessarily the path to get there. It might make sense and be more accessible to allow Python, R, and Swift entries, rather than limiting the results.

JavaScript popped up because there seemed to be quite a few libraries being developed for Node.js. That may be premature for now.

What do you mean by hiring?

I'm evaluating if the business I am in could benefit from running these types of competitions. In the Quantopian contest, the entries are evaluated over 50+ different criteria. I would want to hire one or more individuals who could help us establish parameters for a competition, come up with scoring criteria, filter out people gaming the data, etc. I'm not entirely sure what skillset I should be looking for in order to do that.
 

ChrisV

Legendary Contributor
EPIC CONTRIBUTOR
Read Fastlane!
Read Unscripted!
Speedway Pass
User Power
Value/Post Ratio
225%
May 10, 2015
3,141
7,055
Islands of Calleja
JavaScript popped up because there seemed to be quite a few libraries being developed for Node.js. That may be premature for now.
Oh yea, node.js is pretty sweet for Data Visualization. I stand corrected.

But I still think it's uses are pretty niche.

I think maybe I misunderstood the question. So you want to run competitions and find Data Scientists judge the results?
 
Dislike ads? Remove them and support the forum: Subscribe to Fastlane Insiders.

ZeroToOne

Contributor
Read Fastlane!
Speedway Pass
User Power
Value/Post Ratio
157%
May 6, 2017
44
69
KS
Apologies... I've probably been a little too vague on the overall idea. I am building a platform that will allow me to build and run different AI/ML scripts, similar to the way Quantopian, QuantConnect, or Kaggle have set up their contest environments. Kaggle appears to be mostly longer term, team-oriented contests (?), which is interesting, but different from what I was originally thinking. That is probably a separate discussion. My goals are to:

  • Run competitions so I can validate that my own system functions properly. So in that regard, I'm using the competitions simply as a way to beta test my platform.
  • Eventually attract Data Scientists to the platform as participants/users. By running successful contests, I can build rapport and target the exact type of person I would want to use the platform when it goes into production.
I think maybe I misunderstood the question. So you want to run competitions and find Data Scientists judge the results?

Partially yes. I would want to hire one or more individuals who can give me input on building the platform, assistance on developing/running the contest, and also help with how to judge the results (preferably algorithmically).

Let me know if that makes sense.
 

Saavedra

Contributor
User Power
Value/Post Ratio
92%
Aug 4, 2013
84
77
Apologies... I've probably been a little too vague on the overall idea. I am building a platform that will allow me to build and run different AI/ML scripts, similar to the way Quantopian, QuantConnect, or Kaggle have set up their contest environments. Kaggle appears to be mostly longer term, team-oriented contests (?), which is interesting, but different from what I was originally thinking. That is probably a separate discussion. My goals are to:

  • Run competitions so I can validate that my own system functions properly. So in that regard, I'm using the competitions simply as a way to beta test my platform.
  • Eventually attract Data Scientists to the platform as participants/users. By running successful contests, I can build rapport and target the exact type of person I would want to use the platform when it goes into production.


Partially yes. I would want to hire one or more individuals who can give me input on building the platform, assistance on developing/running the contest, and also help with how to judge the results (preferably algorithmically).

Let me know if that makes sense.


How did it go?
 

Post New Topic

Please SEARCH before posting.
Please select the BEST category.

Post new topic

Guest post submissions offered HERE.

New Topics

Fastlane Insiders

View the forum AD FREE.
Private, unindexed content
Detailed process/execution threads
Ideas needing execution, more!

Join Fastlane Insiders.

Top