Top Banner
Lund 18. December 2008 Institut for Informatik & Matematisk Modellering
12

Institut for Informatik & Matematisk Modellering

Jan 08, 2016

Download

Documents

Nubia

Institut for Informatik & Matematisk Modellering. Forskningsområder (10 Sektioner). Mathematical modelling: Scientific Computing: Per Christian Hansen Mathematical Statistics: Per Bruun Brockhoff Intelligent Signal Processing: Lars Kai Hansen - PowerPoint PPT Presentation
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

Institut for

Informatik & Matematisk Modellering

Page 2: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

• Mathematical modelling:– Scientific Computing: Per Christian Hansen– Mathematical Statistics: Per Bruun Brockhoff– Intelligent Signal Processing: Lars Kai Hansen– Image Analysis & Computer Graphics: Rasmus Larsen

• Computer Science:– Software Engineering: Hubert Baumeister– System Security: Robin Sharp– Language Based Technology: Hanne Riis Nielson– Algorithms and Logic: Paul Fischer– Embedded Systems: Jan Madsen– System-on-Chip: Jens Sparsø

Forskningsområder (10 Sektioner)

Page 3: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

1. Statistics Section (produced 3 Ph.D.s/year in 10 years)

2. Image analysis Section (produced 2-3 Ph.D.s/year in 10 years)

3. Intelligent Signal Processing Section (ISP)(produced 3-4 Ph.D.s/year in 10 years)

Statistics@DTU Informatics

Roughly: 100 Ph.D.s in recent 10 years

Page 4: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

The Data, Users, KnowledgeData, Users, Knowledge Vision

Lars Kai HansenProfessor, Sektionsleder

Cognitive SystemsSection(Potential new name)Integrating Signal Processing, Machine Learning, Cognitive Engineering for the Digital Economy

Page 5: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

• Signal processing – Typical projects: Hearing aids,

audio search– Generalized cocktail party

problem (ICA)• Machine Learning

– Adaptive systems, automated text analysis, pattern classification,

• Systems Neuroscience– Mental State decoding– Meta-analysis/neuroinformatics

• Datamining– Search engines, meta-analysis– Wikipedia research

• Cognitive science– Search engines– Psycho-physics e.g. audio-

visual integration– Computational cognitive

modelling

• Digital economy– Prof Ingemar Cox, UCL adjunct

prof – Social networks, Web 2.0, User

generated content• Mobile services

– Communication, mobile web apps, Web 2.0

– Cell phone as sensor system

• Human computer interfaces– Intelligent text entry

• Digital media semantics– Latent semantics, automated

text analysis, meta-data analysis

Components of the new unit

Page 6: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

Image Analysis Section

• Many Ph.D. students • Many externally funded research projects• Strong on innovation (E.g. Videometer)• Methodological Profile

• Multivariate Statistics• Pattern recognition• Supervised Learning• Shape Analysis

Rasmus Larsen, Professor Sektionsleder

Page 7: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

Statistics SectionPer Bruun Brockhoff, Professor Sektionsleder

STAT InformaticsResearch VIP (1/3 2008) 7.5 48.5VIP for teaching (faculty) 7.5 52.5Refereed articles 03-07 141 372Conference papers 03-07 76 548PhD students (Oct. 1st, 2008) 22 72Number of PhD grad. 03-07 15 72MSc theses 2007 19 118MSc theses 2006 16 113External funding DKKmill. 06-08 26,6 85,2

Some recent key figures: (Self evaluation report, Dec 2008)

Page 8: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

Statistics SectionMethodological Profile

•Dynamics • Time Series Analysis and Forecasting• Modeling of Dynamics Stochastic Systems• Stochastic Control Theory • Applied Probability with Stochastic Processes

•Industrial and Applied Statistics• Statistical design and analysis of experiments• Statistical Process Control• Sensometrics and Chemometrics• Environmetrics

Page 9: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

Statistics SectionMethodological Profile

• Examples:• Climatology• Energy (Wind)• Pharmaceutics• Finance• Food science • Environmental research• Marine biology• Microbiology

• Characteristics:• Many externally funded projects• Many Ph.D. students• Broadly engaged outside the mathematical statistical area

Page 10: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

Statistics SectionExample: Food Science

Formally engaged in:

Visual analysis!Free open-source software developed by MATFORSK and IMM, DTU for analysis of sensory profile data and panel performance monitoring.

Partners:Denmark: (Supported by DFFE) KU Life

DTU-Aqua Danish Meat Research InstituteChr. Hansen A/SChewTech A/SBang & Olufsen

Norway: (Supported by Norwegian Research Council)

Matforsk, Ås, Norway Tine, Gilde, Arcus AS, Hennig Olsen Is, Fiskeriforskning

Page 11: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

ISCC

• 'DTU Informatics Statistical Consulting Center (ISCC)’

• Established in 2004.

• ISCC offers statistical consulting to students, Ph.D. students and staff at DTU and to external individuals, institutions and companies.

Klaus K. Andersen

Page 12: Institut for Informatik & Matematisk Modellering

Lund 18. December 2008

• Has the construction of ‘Øresundsbroen’ affected the stock of glass eel?

ISCCHenrik Spliid