Maurice Dixon

Experience with data mining for the anaerobic wastewater treatment process

Abstract Anaerobic digestion provides an effective way of disposing organic material in wastewater. The EU-funded TELEMAC project aims at improving the reliability and efficiency of monitoring and control of this type of wastewater treatment plant. One of its special features is the idea of a telecontrol centre which monitors multiple, geographically distributed plants remotely, acts as a centre of expertise, and brings together the expertise of a network of remote experts. Data mining has been identified as a potentially useful contributing technology. Sensor data is now becoming available for some pilot, laboratory scale, and industrial sized digesters. This paper presents the directions of work and emerging results of data mining. Particular themes considered here include: • experience gained in the data mining exercise; • the use of confidence and prediction intervals; • prospects for generalisation over different sizes and types of anaerobic digester; • relationship to the overall supervision system developed in the project.

Non-parametric extraction of implied asset price distributions

We present a fully non-parametric method for extracting risk neutral densities (RNDs) from observed option prices. The aim is to obtain a continuous, smooth, monotonic, and convex pricing function that is twice differentiable. Thus, irregularities such as negative probabilities that afflict many existing RND estimation techniques are reduced. Our method employs neural networks to obtain a smoothed pricing function, and a central finite difference approximation to the second derivative to extract the required gradients.

Object-Oriented Graceful Evolution Monitors

Software development teams are required to produce applications that are enmeshed with contributory systems over which the team has no control. This highlights the need for an approach that allows the developed application to evolve gracefully with changes in the contributory systems. This work proposes an approach to graceful evolution which is appropriate for an object-oriented rapid application development environment. The approach combines elements of Risk Analysis (Baskerville and Stage, 1996) and of the Goal Based Requirements Analysis Method (Anton, 1997), with the perspective given by considering Dynamic Inconsistency (Lamsweerde, Letier and Ponsard, 1997). The approach investigates assumptions made about requirements; the obstacles to those assumptions are then identified. The obstacles are assessed with respect to their impact on the running system and the decision is made to resolve, monitor or ignore the obstacle. The assessment provides (both directly and through the monitoring logs) guidance to the software development teams of the type of corrective action needed. This work demonstrates the approach for a synthetic example drawn from experience in the telecommunications industry for which the enmeshed system was a legacy system.

Grids in control of anaerobic wastewater treatment plants: Leveraging the knowledge

Abstract A potential application of Grids systems is to anaerobic wastewater treatment plants (WWTP). This work extends previous simulation and data mining with a laboratory scale anaerobic digester, and shows that extending to an industrial scale plant gives valid and useful results for monitoring and control. The idea of a telecontrol centre for managing a group of plants is gaining acceptance. We investigate what barriers arise in applying data mining across multiple plants, in a Grids environment. One issue is heterogeneities, implying that data mining must be enhanced with the use of metadata. We examine the heterogeneities, link them to the model of Grids, and investigate their implications for metadata representation.

Trust obstacle mitigation for database systems

This paper introduces the Trust Obstacle Mitigation Model (TOMM), which uses the concept of trust assumptions to derive security obstacles, and the concept of misuse cases to model obstacles. The TOMM allows a development team to anticipate malicious behaviour with respect to the operational database application and to document a priori how this malicious behaviour should be mitigated.

Distributed Data Mining and Knowledge Management with Networks of Sensor Arrays

Environmental pollution control relies heavily on human expert judgment sup- ported by historical data and scientific models. Telemonitoring, by networks of heterogeneous sensor arrays, provides the opportunity for data mining models to be constructed from the historical data to supplement human expertise. This paper reports some progress made in the TELEMAC project by data mining. TELEMAC is concerned with enhancing the efficacy of anaerobic digestion in potentially unstable digesters. In the laboratory using full instrumentation it is possible to derive a good description of the digester state. With data mining it is possible to identify some constraints on sensor choice. This paper examines this data mining work from the perspective of a three layer Grid architecture to see what implications and requirements arise that could benefit the exercise of expert judgment. After placing the specific TELEMAC situation in a generic Grids context, we present a classification approach to attributes for metadata and indicate some examples of model resource discovery.