-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy path2017-09-01-euroscipy.html
67 lines (62 loc) · 5.89 KB
/
2017-09-01-euroscipy.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<link rel="alternate"
type="application/rss+xml"
href="https://bastibe.de/rss.xml"
title="RSS feed for https://bastibe.de/">
<title>EuroScipy 2017</title>
<meta name="author" content="Bastian Bechtold">
<meta name="referrer" content="no-referrer">
<link href= "static/style.css" rel="stylesheet" type="text/css" />
<link rel="icon" href="static/favicon.ico">
<link rel="apple-touch-icon-precomposed" href="static/favicon-152.png">
<link rel="msapplication-TitleImage" href="static/favicon-144.png">
<link rel="msapplication-TitleColor" href="#0141ff">
<script src="static/katex.min.js"></script>
<script src="static/auto-render.min.js"></script>
<script src="static/lightbox.js"></script>
<link rel="stylesheet" href="static/katex.min.css">
<script>document.addEventListener("DOMContentLoaded", function() { renderMathInElement(document.body); });</script>
<meta http-equiv="content-type" content="application/xhtml+xml; charset=UTF-8">
<meta name="viewport" content="initial-scale=1,width=device-width,minimum-scale=1"></head>
<body>
<div id="preamble" class="status"><div class="header">
<a href="https://bastibe.de">Basti's Scratchpad on the Internet</a>
<div class="sitelinks">
<a href="https://github.com/bastibe">Github</a> | <a href="https://bastibe.de/projects.html">Projects</a> | <a href="https://bastibe.de/uses.html">Uses</a> | <a href="https://bastibe.de/reviews.html">Reviews</a> | <a href="https://bastibe.de/about.html">About</a>
</div>
</div></div>
<div id="content">
<div class="post-date">01 Sep 2017</div><h1 class="post-title"><a href="https://bastibe.de/2017-09-01-euroscipy.html">EuroScipy 2017</a></h1>
<p>
The second conference I attended this year was the <a href="https://www.euroscipy.org/2017/">EuroScipy 2017</a> in Erlangen. I gave a <a href="https://www.youtube.com/watch?v=mc8ru37dwf8">Talk about Audio in Python</a> and a <a href="https://youtu.be/qTgk2DUM6G0?t=10m15s">Lightning Talk about my Python/Matlab bridge</a>.
</p>
<p>
My most striking impression of EuroScipy is that every person I talked to was working on something interesting, and could talk about his/her topic clearly and with enthusiasm. This mirrors my feelings from last year's Chaos Communication Congress, where the short scientific section stood out for its clarity and passion. I also enjoyed the fact that attendees were international and diverse, and exuded a heart-warming sense of community.
</p>
<p>
Even though each scientific discipline has their own data sets, features, and models, everyone seemed to use a common set of methods (statistics, signal processing, machine learning) for working with that data. And, absolutely everybody used <a href="http://jupyter.org/">Jupyter Notebooks</a> for tutorials and teaching, and almost all data analyses were done in <a href="http://pandas.pydata.org/">Pandas</a>. This is particularly heartening since these technologies are geared towards reproducible research and open data.
</p>
<p>
The hot topic of the conference clearly was machine learning and neural networks. However, the current confusion of competing frameworks and network architectures does not seem to be a good long-term solution. I hope that this ecosystem will eventually reach its NumPy moment, and collapse into a single, unified package. Then, neural networks might find their place as just another machine learning method with a few reusable parametrized prototype implementations in scikit-learn. Tools like <a href="https://keras.io/">Keras</a> look like good steps towards this goal.
</p>
<p>
Finally, there was a lot of talk about “the reproducibility crisis” in science, and possible steps to improve the scientific process. In particular, I learned that it is absolutely necessary to not look at your test data before the final evaluation, to avoid overfitting your brain. You need to split your data into a development set for training, a validation set for parameter tuning, and a totally separate evaluation set for the final evaluation. In a similar way, it is important to state your hypotheses <i>in writing</i> before you test them, to avoid “HARKing” (Hypothesis After the Results are Known; aka “Noise Mining”, “P-Hacking”, or “Procedural Overfitting”). I dearly hope that <a href="https://cos.io/rr/">Registered Reports</a> will catch on, and absolve us from these all too human biases.
</p>
<p>
In conclusion, EuroScipy 2017 was a ton of fun, and educational in many ways that I did not expect. If you are a scientific programmer, or if you maintain a scientific Python module, or if you are plain interested in scientific Python, I can highly recommend going to EuroScipy next year.
</p>
<div class="taglist"><a href="https://bastibe.de/tags.html">Tags</a>: <a href="https://bastibe.de/tag-event.html">event</a> </div>
<div id="comments"><script async src="https://talk.hyvor.com/embed/embed.js" type="module"></script>
<hyvor-talk-comments id="hyvorcomments" website-id="3390" page-id=""></hyvor-talk-comments>
<script type="text/javascript">
document.getElementById("hyvorcomments").setAttribute("page-id", location.pathname);
</script></div></div>
<div id="postamble" class="status"><div id="archive">
<a href="https://bastibe.de/archive.html">Other posts</a>
</div>
<center><a rel="license" href="https://creativecommons.org/licenses/by-sa/3.0/"><img alt="Creative Commons License" style="border-width:0" src="https://i.creativecommons.org/l/by-sa/3.0/88x31.png" /></a><br /><span xmlns:dct="https://purl.org/dc/terms/" href="https://purl.org/dc/dcmitype/Text" property="dct:title" rel="dct:type">bastibe.de</span> by <a xmlns:cc="https://creativecommons.org/ns#" href="https://bastibe.de" property="cc:attributionName" rel="cc:attributionURL">Bastian Bechtold</a> is licensed under a <a rel="license" href="https://creativecommons.org/licenses/by-sa/3.0/">Creative Commons Attribution-ShareAlike 3.0 Unported License</a>.</center></div>
</body>
</html>