Using HIT to deliver integrated care for the frail elderly in the UK: current barriers and future challenges

In this paper we briefly describe the results of a 3 year project examining the use of Health Information Technologies (e.g., electronic patient record systems) to deliver integrated care. In particular, we focus on one group of patient (the frail elderly) and efforts to design an e-health supported healthcare pathway (the frail elderly pathway – FEP). The aim of FEP is to bring together clinicians and staff from health and social care and allow them to share patient information. Our findings show that progress in delivering a fully-supported and working FEP has been slow, not least because of the difficulties experienced by healthcare staff in using current IT systems. In addition, there are many strategic and technical issues which remain unresolved (e.g., systems interoperability).

A novel satellite mission concept for upper air water vapour, aerosol and cloud observations using integrated path differential absorption LiDAR limb sounding

We propose a new satellite mission to deliver high quality measurements of upper air water vapour. The concept centres around a LiDAR in limb sounding by occultation geometry, designed to operate as a very long path system for differential absorption measurements. We present a preliminary performance analysis with a system sized to send 75 mJ pulses at 25 Hz at four wavelengths close to 935 nm, to up to 5 microsatellites in a counter-rotating orbit, carrying retroreflectors characterized by a reflected beam divergence of roughly twice the emitted laser beam divergence of 15 µrad. This provides water vapour profiles with a vertical sampling of 110 m; preliminary calculations suggest that the system could detect concentrations of less than 5 ppm. A secondary payload of a fairly conventional medium resolution multispectral radiometer allows wide-swath cloud and aerosol imaging. The total weight and power of the system are estimated at 3 tons and 2,700 W respectively. This novel concept presents significant challenges, including the performance of the lasers in space, the tracking between the main spacecraft and the retroreflectors, the refractive effects of turbulence, and the design of the telescopes to achieve a high signal-to-noise ratio for the high precision measurements. The mission concept was conceived at the Alpbach Summer School 2010.

Modular non-computational-connectionist-hybrid neural network approach to robotic systems

Spiking neural networks are usually limited in their applications due to their complex mathematical models and the lack of intuitive learning algorithms. In this paper, a simpler, novel neural network derived from a leaky integrate and fire neuron model, the ‘cavalcade’ neuron, is presented. A simulation for the neural network has been developed and two basic learning algorithms implemented within the environment. These algorithms successfully learn some basic temporal and instantaneous problems. Inspiration for neural network structures from these experiments are then taken and applied to process sensor information so as to successfully control a mobile robot.

Understanding lake and catchment history as a tool for integrated lake management

Sustainable lake management for nutrient-enriched lakes must be underpinned by an understanding of both the functioning of the lake, and the origins of changes in nutrient loading from the catchment. To date, limnologists have tended to focus on studying the impact of nutrient enrichment on the lake biota, and the dynamics of nutrient cycling between the water column, biota and sediments within the lake. Relatively less attention has been paid to understanding the specific origins of nutrient loading from the catchment and nutrient transport pathways linking the lake to its catchment. As such, when devising catchment management strategies to reduce nutrient loading on enriched lakes, assumptions have been made regarding the relative significance of non-point versus point sources in the catchment. These are not always supported by research conducted on catchment nutrient dynamics in other fields of freshwater science. Studies on nutrient enrichment in lakes need to take account of the history of catchment use and management specific to each lake in order to devise targeted and sustainable management strategies to reduce nutrient loading to enriched lakes. Here a modelling approach which allows quantification of the relative contribution of nutrients from each specific point and non-point catchment source over the course of catchment history is presented. The approach has been applied to three contrasting catchments in the U.K. for the period 1931 to present. These are the catchment of Slapton Ley in south Devon, the River Esk in Cumbria and the Deben Estuary in Suffolk. Each catchment showed marked variations in the nature and intensity of land use and management. The model output quantifies the relative importance of point source versus non-point livestock and land use sources in each of the catchments, and demonstrates the necessity for an understanding of site-specific catchment history in devising suitable management strategies for the reduction of nutrient loading on enriched lakes.

Suspended and bed load sediment transport dynamics in two lowland UK streams—storm integrated monitoring

A great deal of work recently has focused on suspended and bedload sediment transport, driven primarily by interest in contaminant transfer. However, uncertainties regarding the role of storm events, macrophyte beds and interactions between the two phases of sediment still exist. This paper compares two study sites within the same catchment whose geology varies significantly. The differences in hydrology, suspended sediment (SS) transport and bed load transport that this causes are examined. In addition, a method to predict the mobilization of different size fractions of sediment during given flows is investigated using critical entrainment thresholds.

PMCRI: a parallel modular classification rule induction framework

In a world where massive amounts of data are recorded on a large scale we need data mining technologies to gain knowledge from the data in a reasonable time. The Top Down Induction of Decision Trees (TDIDT) algorithm is a very widely used technology to predict the classification of newly recorded data. However alternative technologies have been derived that often produce better rules but do not scale well on large datasets. Such an alternative to TDIDT is the PrismTCS algorithm. PrismTCS performs particularly well on noisy data but does not scale well on large datasets. In this paper we introduce Prism and investigate its scaling behaviour. We describe how we improved the scalability of the serial version of Prism and investigate its limitations. We then describe our work to overcome these limitations by developing a framework to parallelise algorithms of the Prism family and similar algorithms. We also present the scale up results of a first prototype implementation.

Parallel induction of modular classification rules

The Distributed Rule Induction (DRI) project at the University of Portsmouth is concerned with distributed data mining algorithms for automatically generating rules of all kinds. In this paper we present a system architecture and its implementation for inducing modular classification rules in parallel in a local area network using a distributed blackboard system. We present initial results of a prototype implementation based on the Prism algorithm.

Towards a computationally efficient approach to modular classification rule induction

Induction of classification rules is one of the most important technologies in data mining. Most of the work in this field has concentrated on the Top Down Induction of Decision Trees (TDIDT) approach. However, alternative approaches have been developed such as the Prism algorithm for inducing modular rules. Prism often produces qualitatively better rules than TDIDT but suffers from higher computational requirements. We investigate approaches that have been developed to minimize the computational requirements of TDIDT, in order to find analogous approaches that could reduce the computational requirements of Prism.

J-PMCRI: a methodology for inducing pre-pruned modular classification rules

Inducing rules from very large datasets is one of the most challenging areas in data mining. Several approaches exist to scaling up classification rule induction to large datasets, namely data reduction and the parallelisation of classification rule induction algorithms. In the area of parallelisation of classification rule induction algorithms most of the work has been concentrated on the Top Down Induction of Decision Trees (TDIDT), also known as the ‘divide and conquer’ approach. However powerful alternative algorithms exist that induce modular rules. Most of these alternative algorithms follow the ‘separate and conquer’ approach of inducing rules, but very little work has been done to make the ‘separate and conquer’ approach scale better on large training data. This paper examines the potential of the recently developed blackboard based J-PMCRI methodology for parallelising modular classification rule induction algorithms that follow the ‘separate and conquer’ approach. A concrete implementation of the methodology is evaluated empirically on very large datasets.

Induction of modular classification rules: using Jmax-pruning

The Prism family of algorithms induces modular classification rules which, in contrast to decision tree induction algorithms, do not necessarily fit together into a decision tree structure. Classifiers induced by Prism algorithms achieve a comparable accuracy compared with decision trees and in some cases even outperform decision trees. Both kinds of algorithms tend to overfit on large and noisy datasets and this has led to the development of pruning methods. Pruning methods use various metrics to truncate decision trees or to eliminate whole rules or single rule terms from a Prism rule set. For decision trees many pre-pruning and postpruning methods exist, however for Prism algorithms only one pre-pruning method has been developed, J-pruning. Recent work with Prism algorithms examined J-pruning in the context of very large datasets and found that the current method does not use its full potential. This paper revisits the J-pruning method for the Prism family of algorithms and develops a new pruning method Jmax-pruning, discusses it in theoretical terms and evaluates it empirically.