Munich Datageeks - June Edition

The summer is back in Munich and so are the Datageeks. This time we will get together at the offices of our first time sponsor Codecentric (https://www.codecentric.de). Special thanks to them in advance.

The format is as always:

• 2 presentations (each ca. 30-40 min incl. discussion)
• time for networking + food + drinks before, in between and especially after the presentations
• talks are held in English

The line up:

1. Mathias Fuchs
   Topic: "How to measure the performance of a machine learning algorithm"

Talk Description Machine learning has seen an impressing boom in the past years. In practice, every learning algorithm has to prove its worth by better classification/prediction performance than previous algorithms. However, obvious performance measures such as the misclassification rate and so on suffer a high variability across data-sets or even re-sampling iterations. Thus, it is a challenging problem to assess the performance reliably, and its importance has been given surprisingly little attention in literature and practice. I will talk about the usual performance measures such as misclassification rate, AUC or mean squared error etc. I will explain cross-validation and its pitfalls. Finally, I will show you a new method to estimate the variability of the performance estimators in a general and statistically correct way.

Speaker Bio Mathias studied mathematics in Munich and Marseille. After his PhD, he went across numerical mathematics and bioinformatics, to arrive in biostatistics and machine learning where he feels at home now. Currently, he is Postdoc at the medical faculty of LMU, enthusiastic about the fascinating challenges in big-data-statistics and current machine learning.

2. Jörg Blumtritt
   Topic: "Mobile data: under the hood. Generating, collecting, and processing smartphone data."

Talk Description - The phone's sensors: from geolocation to magnetic flux (& getting access with various apps with their advantages and problems)

• Noise and artefacts
• Battery issues
• Complex event processing
• Tracking behavior
• Privacy: How to track spooky things and how to make people aware of what you are doing.

Speaker Bio - studied statistics, mathematics, political sciences;

• has worked in behavioral science research in Max-Planck-Gesellschaft (nonverbal behavior, origins of xenophobia), for various TV networks, publishers (head of research with Hubert Burda Media), agencies (managing director MediaCom Germany),
• founded Datarella in 2012 with focus on computational social science
• blogging at http://datarella.com/blog, http://beautifuldata.net, http://slow-media.net