Waikato University crest

Department of
Computer Science
Tari Rorohiko

Computing and Mathematical Sciences

Upcoming Seminars


Feature Engineering for Handling Spatial and Spatio-Temporal Forecasting

Luis Torgo
LIAAD-INESC Tec / University of Porto, Portugal
Tuesday 29 September 2015
11:00 am
In this talk we will describe an approach to spatial and spatio-temporal forecasting tasks that is based on simple feature engineering techniques. This approach is focused on creating new features with the goal of providing the models with relevant information on the spatial and spatio-temporal dynamics present in the available data. We describe this approach that allows the usage of widely available regression tools on this type of tasks. We also present two concrete applications of this simple method: (i) an application to image repair; and (ii) and application on wind speed forecasting in the context of wind power generation. Extensive experimental results show the competitiveness of this simple approach.

Luis is visiting the Department from now until Christmas. He is currently based in the Math Visitors' Office on the third floor of G Block.


My Journey to Data Mining

Hans-Peter Kriegel
Informatics at Ludwig-Maximilians-Universitaet Munich, Germany
Tuesday 6 October 2015
11:00 am
In this talk, the speaker will describe how he came as a database researcher (which he has not given up) to the field of data mining. The talk will present the density-based principle and its application to clustering and outlier detection, including high dimensional data. At the end of the talk we want to give a short outlook at the art of runtime evaluation. Are we comparing algorithms or implementations?

Biography: Hans-Peter Kriegel has been a Professor of Informatics at Ludwig-Maximilians-Universitaet Munich, Germany since 1991. He has published over a wide range of data mining topics including clustering, outlier detection and high-dimensional data analysis. His original and still active research area ist database systems, in particular query processing, index and access processing, management of spatial and uncertain data. In 2009 the Association for Computing Machinery (ACM) elected Professor Kriegel an ACM Fellow for his contributions to knowledge discovery and data mining, similarity search, spatial data management, and access-methods for high-dimensional data. He received the 2013 IEEE ICDM Research Contributions Award for his influential contributions to the field of data mining. This year he received the 2015 SIGKDD Innovation Award for his outstanding technical contributions to the KDD field. So far, his more than 450 publications have been cited more than 35,000 times according to Google Scholar. Microsoft Academic Search currently ranks him Number 5 for field rating in the field of data mining.

Professor Kriegel's ground-breaking contribution to the data mining field was his paper at the 1996 KDD Conference titled "A density-based algorithm for discovering clusters in large spatial databases with noise" (DBSCAN) with co-authors Martin Ester, Joerg Sander and Xiaowei Xu, with more than 7,000 citations in Google Scholar. This paper received the 2014 SIGKDD Test of Time Award. It also led to other density-based approaches such as OPTICS (Ordering Points To Identify the Clustering Structure, SIGMOD 1999) and LOF (Identifying density-based Local Outliers, SIGMOD 2000), both with more than 2,000 citations. His more recent work on clustering as well as on outlier detection in high-dimensional data was presented in numerous tutorials at IEEE ICDM and ACM SIGKDD, among other venues. This research has been done together with his team members Peer Kroeger, Erich Schubert and Arthur Zimek.

More recently, Professor Kriegel has been applying some of his work on clustering and outlier detection in interdisciplinary cooperations with other sciences such as archeology, biology, engineering and medical science as well as in industrial collaborations.


Events Index