These are chat archives for FreeCodeCamp/DataScience

6th
Sep 2017
Eric Leung
@erictleung
Sep 06 2017 01:14
Interesting. Uber's machine learning creation for ML-as-a-service system https://eng.uber.com/michelangelo/ in order to democratize machine learning analysis, covering end-to-end ML workflows (managing data to making predictions and monitoring its performance).
However, I'm more interested in this paper they linked to, Hidden Technical Debt in Machine Learning Systems (PDF), which I think was summed up as the "anti-pattens for machine learning." I haven't read it yet, but might be useful in large scale machine learning projects.
Oh and that last paper was written by researchers at Google, so they definitely know a thing or two about large scale projects.
evaristoc
@evaristoc
Sep 06 2017 20:28

People,

For those interested in information retrieval and topics regarding recommender systems please contact me. I have in mind a pipeline of several projects to work on.

In principle we are going to work with python but there is a chance that we could rewrite part of the code into Javascript in the future. As a minimum you need to have a good command of Algebra and Stats. Very nice if you are into Python and seriously learning Javascript. If you are also skilled in web development, you are MORE than welcome. If you are also a ML/DS student or graduate or something in Data Engineering, MUCH BETTER.

If you don't know what a recommender system is, just some examples:

  • Amazon
  • Expedia
  • Netflix
  • Spotify, SoundCloud, Pandora

We will be implementing elementary techniques for now as we are just working on a PoC. NO Big Data for now, I am afraid.

As an example of the kind of projects please check an unfinished demo of one of the projects: https://evaristoc.github.io/fCC_R3c/

Matthew Barlowe
@mcbarlowe
Sep 06 2017 22:35
I'd love to help out I'm a data analyst student I know python pretty well. I only know a little JavaScript but I've been wanting to learn. I don't know how much help I'll be because I'm still new at this but I'll at least try