tnarik / training-odsc_europe_2017

Training material compiled from ODSC Europe 2017 (London)

ODSC Europe 2017

Description

This is a bundle compiling information (and snapshots of extenal repositories) related to the ODSC Europe sessions of 2017, held October 12-14 in London).

Not all sessions included code or were all that interesting, so there might be some not corresponding with a folder.

Sessions

The intention is creating a folder per session attended, including slides and links when available. If no content is available for a certain session, no content will be provided. The sessions are sorted by a two digit prefix.

1. Algorithmic Trading, by Yves Hilpisch:

   • Code from https://gist.github.com/yhilpisch/35a18382cfdd850c55fe971ae54cf752 or https://goo.gl/LAQ8Ze

2. Deep Learning with TensorFlow for absolute beginners, by Kaz Sato and Matthias Feys:

   • Presentation at https://docs.google.com/presentation/d/1OzUurRlaHHr_z9eiRFzlPWrr64Jtx1kthrhyuK1Bm8s/edit?usp=sharing
   • This workshop was way too basic and we spent most of the time setting up Google Codelabs accounts and instances. I ended up downloading the dataset from https://opendata.cityofnewyork.us/, instead of using directly from the cloud managed BigQuery dataset. It was easy downloading it locally (it takes 240 MB or 40 MB compressed).

3. Probabilistic programming, by Tomas Wiecki:

   • This was one of a series of sessions in quantitative trading (or Quant Finance), mostly centered around Quantopian for the sessions.
   • The most interesting point - other than the idea of making money :) - was Bayesian statistics applications.

4. Insights from Complex Network Data, by Pablo Suau.

5. Deep Learning for Developers, by Julien Simon.

6. Deep Learning in Keras, by Leonardo de Marchi.

7. Introduction to Algorithmic Trading, by Max Margenot.

8. Introduction to Bayesian Analysis in Python, by Peadar Coyle.

9. Open Containerization Tooling, by Phil Weir.

10. Training, Managing and Deploying Models, by Piotr Migdał: essentially how to use Neptune.

11. Bayesian A/B testing, by Marc Garcia.

12. TDD-ing a Bayesian Classifier, by Robert Hardy: a look at pytest.

Environment

I'm currently using my own environment based on conda https://github.com/tnarik/env-machine_learning.

Notebooks

When present, *-work.ipynb notebooks are the worked upon version of the corresponding notebook. That might include additional comments, path modification, etc.

Moving forward

Apart from pointers and links to libraries, techonologies and concepts (is always good to have an overview of what's out there), I started thinking on a couple of ideas and toy projects (either for practice or to further explore some new territories). Those won't be available here as they deserve their own repositories.