Network on multiScale Information, RePresentatIon and Estimation -- (INSPIRE)

点击翻译按钮获取中文摘要

Today's society is experiencing a revolution in data acquisition and storage. New forms of data with unprecedented levels of heterogeneity and complexity are collected routinely in areas such as medical imaging, biochemistry, and bioinformatics. As the success of data acquisition continues, and the collection of data become easier to implement, data formats become ever more diverse and the number of outstanding problems in developing new tools for understanding the world we live in, is increasing rather than diminishing. To be able to rationally make inferences from data, our understanding of its mechanism must be made quantitative and precise. With the increasing level of difficulty in this task inherent in qualifying the generation of more complex phenomena, expertise must be brought together from different areas of science, notably signal processing, statistics and mathematics, to tackle the problems of modelling, representation and analysis. Despite this universally acknowledged fact, developments in the aforementioned areas, are often divergent and unsynchronized. This application therefore proposes to establish a network of researchers, working on problems of structured data representation and inference using multiscale and related methods. The network seeks to build on existing strong UK groups that have been working mainly independently, by connecting their expertise and developments. By creating a virtual centre of excellence to share information and collaborate, and by holding regular meetings, we seek to ensure a steady and strong flow of information between the proposed network nodes. This will have an impact beyond the initially proposed membership: throughout the lifespan of the network we intend to work towards its expansion, and seek to form an inclusive virtual centre of excellence, bridging the gap between disciplines. Furthermore, the establishment of a network in this area with a strong multidisciplinary component, will feed into areas that use methods for analysis of structured forms of data, where examples include such disparate fields such as medical imaging and finance. By including a natural forum for multidisciplinary interaction and training of graduate students from all the related areas, we seek to foster a spirit of multidisciplinary interaction in the next generation of researchers, and secure a lasting impact of the network on the scientific community. Ours is an information age, and only by investing in the development of tools to collect and analyse information, can we hope to make headway into the scientific problems of tomorrow. This development can only be done effectively by using the expertise from all the relevant disciplines, and so it is necessary to connect the outlined areas of research, as proposed by this network.