Munich Datageeks - June Edition

It is time for our next Meetup. MaibornWolff will host for the second time.

Format:

• 2 talks (each ca. 40 min incl. discussion)

• Time for networking + food + drinks before, in between and after the presentations

• Talks are held in English

• We will be taking photos and/or film footage at the event. These will be used to share news about our meetups, and to publicize upcoming events.

• We will provide a list of registered usernames to the sponsor to check at the entry.

• Please bring photo identification to be checked at the entrance.

Luca Massaron - Kaggle PetFinder Competition: Machine Learning For Puppies

Abstract:

PetFinder.my has been Malaysia’s leading animal welfare platform since 2008, building in the meantime a rich database of more than 150,000 animals, completed with their history and photos. Recently it held a competition on Kaggle, requiring participants to develop algorithms to predict the adoptability of pets - specifically, how quickly a pet is going to be adopted. The problem mixed tabular data, textual information and varied images, requiring finding an optimal mix of preprocessing and learning techniques in order to achieve a performing result. Our team, comprising 3 Kaggle masters and 3 Kaggle experts from 3 three different countries, achieved the 8th rank in the competition and thereafter open-sourced our solution. In the talk, we illustrate the organizational and technical challenges (with their solutions) we had to face in order to achieve such a result.

Bio:

Luca Massaron is a data scientist and a research director specialized in multivariate statistical analysis, machine learning and customer insight with over fifteen years of experience in solving real world problems and in generating value for stakeholders by applying reasoning, statistics, data mining and algorithms. From being a pioneer of Web audience analysis in Italy to achieving the worldwide rank of top ten data scientist at competitions held by kaggle.com, he has always been passionate about everything regarding data and analysis and about demonstrating the potentiality of data-driven knowledge discovery to both experts and non-experts. He is the author of books on Data Science, Machine Learning, Deep Learning and AI. He is also a Google Developer Expert in Machine Learning.

Fabian Hertwig - Predicting New York City Taxi demand: spatio-temporal Time Series Forecasting

Abstract:

Time series forecasting has always been an important field in machine learning and statistics, as it helps us to make decisions about the future. A special field is spatio-temporal forecasting, where predictions are not only made on the temporal dimension, but also on a regional dimension.
In this session, we will present a demonstration project to predict taxi demand in Manhattan, NYC for the next hour. We’ll show some of the basic principles of time series forecasting and compare different models suited for the spatio-temporal use case. We will take a closer look at the principles of models like long short-term memory networks and temporal convolutional networks. We will show that these models decrease the prediction error by 40% as compared to a simple baseline model, which predicts the same demand as in the last hour.

Bio:

I am a Data Scientist at MaibornWolff in Munich. I fell in love with problem solving in data four years ago when I was a working student and helped an automotive company to improve its processes by analyzing data. Since then, I specialized in Data Science and Deep Learning and have worked on various projects for the last three years.