Conference on Data Science
Due to the corona virus crisis, the conference is cancelled.
The conference aims at gathering researchers from various fields around the topics of Statistical Analysis, Probabilistic Graphical Model, Time Series Analysis, Neural Networks and Feature Selection.
In particular we hope that it will give the opportunity to researchers from different fields but working with similar objects to discuss and share ideas.
We aim at keeping the conference informal so that the focus will be on the understanding of each others topics. Consequently, no proceeding will be published following the conference.
Where and When
3rd August: A one-day lecture on data science with Julia will be held at the department of mathematical sciences, Skjernvej 4A, 9220 Aalborg Ø.
4-6th August: The conference will be held at the University of Aalborg, Rendsburggade 14, 9000 Aalborg. The talks will take place in the seminar room 5.125.
- Andreas Noack - Julia Computing
- Niels Richard Hansen - Department of Mathematical Sciences, Copenhagen University
- Mikke Innes - Julia Computing
- Rasmus Waagepetersen - Department of Mathematical Sciences, Aalborg University
- Zheng-Hua Tan -
Department of Electronic Systems, Aalborg University
Title: Deep Representation Learning for Speech and Multimodal Signal Processing
Abstract: Deep learning has significantly advanced the state-of-the-art in speech and multimodal processing, leading to considerable success in a wide range of real-world applications. The core of deep learning essentially lies in learning high-level representations that facilitate the end learning goals or the downstream tasks. This talk will first present discriminative feature learning based on supervised, weakly supervised or unsupervised learning, with application to speech classification. Besides learning a representation in a new space, deep learning has shown to be a powerful tool for recovering corrupted data in the original space. In the second part of this talk, I will discuss methods for recovering speech signals in noise. Finally, I will consider deep learning for multimodal signal processing for various applications. This involves both joint representation learning and data fusion that are made flexible due to the flexibility of deep structures.
- The workshop is funded by the Carlsberg Foundation.
- Thanks to Carlsberg Foundation, we have access to limited funds to pay for the accommodation of several young researchers. Please ask us only if no other funding is available.
- Lunches and refreshments will be provided by the conference but the dinners are not.
- Christophe A.N. Biscio - firstname.lastname@example.org
- Eduardo Vera-Valdes - email@example.com
To be announced - The conference will start the 4th August on the morning and end the 6th August at 12:00.
The 3rd August a one-day lecture on data science with Julia will be held at the department of mathematical sciences.
The aim of this day is to give an overview of the main algorithms in data science and use them with Julia.
No knowledge of Julia is required to attend the course. An introction to Julia will be given to the participants.
A large part of the day will be devoted to exercices. Thus we recommend the participants to bring their computer.
The code used during the lectures will be freely available under the form of Jupyter Notebooks.
Preliminary Program, 9h -16h
Morning - 1st session: Introduction to Julia
Morning - 2nd session: Data science algorithm presentation (theory)
Afternoon - 1st session: Data science algorithm with Julia
Afternoon - final session: exercices
List of speakers
The workshop is held at the University of Aalborg - Denmark.
Venue to Aalborg
We recommend participants outside of Denmark to come by plane to Aalborg aiport and then take a bus or a taxi to their hotel.
The city is also well connected by train.
Note that google map is working very well in Aalborg so that it is very easy to find your way.
Venue to the Conference
Venue to the one-day lecture on Data Science in Julia
The conference will be held in the center of Aalborg city at walking distance of several hotels. For examples:
We have a limited amount of funds that allows us, on request, to pay hotel expanses for several participants. Please contact the organisers if you are interested.