Genetic Algorithm Search for Predictive Patterns in Multidimensional Time Series

Based on an algorithm for pattern matching in character strings, we implement a pattern matching machine that searches for occurrences of patterns in multidimensional time series. Before the search process takes place, time series are encoded in user-designed alphabets. The patterns, on the other hand, are formulated as regular expressions that are composed of letters from these alphabets and operators. Furthermore, we develop a genetic algorithm to breed patterns that maximize a user-defined fitness function. In an application to financial data, we show that patterns bred to predict high exchange rates volatility in training samples retain statistically significant predictive power in validation samples.

Multidimensional modelling of X-ray spectra for AGN accretion disc outflows

We use a multidimensional Monte Carlo code to compute X-ray spectra for a variety of active galactic nucleus (AGN) disc-wind outflow geometries. We focus on the formation of blueshifted absorption features in the Fe K band and show that line features similar to those which have been reported in observations are often produced for lines of sight through disc-wind geometries. We also discuss the formation of other spectral features in highly ionized outflows. In particular, we show that, for sufficiently high wind densities, moderately strong Fe K emission lines can form and that electron scattering in the flow may cause these lines to develop extended red wings. We illustrate the potential relevance of such models to the interpretation of real X-ray data by comparison with observations of a well-known AGN, Mrk 766. Journal compilation © 2008 RAS.

Multidimensional modelling of X-ray spectra for AGN accretion disc outflows - II

Highly ionized fast accretion disc winds have been suggested as an explanation for a variety of observed absorption and emission features in the X-ray spectra of active galactic nuclei. Simple estimates have suggested that these flows may be massive enough to carry away a significant fraction of the accretion energy and could be involved in creating the link between supermassive black holes and their host galaxies. However, testing these hypotheses, and quantifying the outflow signatures, requires high-quality theoretical spectra for comparison with observations. Here, we describe extensions of our Monte Carlo radiative transfer code that allow us to generate realistic theoretical spectra for a much wider variety of disc wind models than that was possible in our previous work. In particular, we have expanded the range of atomic physics simulated by the code so that L- and M-shell ions can now be included. We have also substantially improved our treatment of both ionization and radiative heating such that we are now able to compute spectra for outflows containing far more diverse plasma conditions. We present example calculations that illustrate the variety of spectral features predicted by parametrized outflow models and demonstrate their applicability to the interpretation of data by comparison with observations of the bright quasar PG1211+143. We find that the major features in the observed 2-10 keV spectrum of this object can be well reproduced by our spectra, confirming that it likely hosts a massive outflow. © 2010 The Authors. Journal compilation © 2010 RAS.

Multidimensional rehabilitation programmes for adult cancer survivors

Background: Multidimensional rehabilitation programmes (MDRPs) have developed in response to the growing number of people living with and surviving cancer. MDRPs comprise a physical component and a psychosocial component. Studies of the effectiveness of these programmes have not been reviewed and synthesised.

Objectives: To conduct a systematic review of studies examining the effectiveness of MDRPs in terms of maintaining or improving the physical and psychosocial well-being of adult cancer survivors.

Search methods: We conducted electronic searches in the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, EMBASE, CINAHL and PsychINFO up to February 2012.

Selection criteria: Selection criteria focused on randomised controlled trials (RCTs) of multidimensional interventions for adult cancer survivors. Interventions had to include a physical component and a psychosocial component and to have been carried out on two or more occasions following completion of primary cancer treatment. Outcomes had to be assessed using validated measures of physical health and psychosocial well-being. Non-English language papers were included.

Data collection and analysis: Pairs of review authors independently selected trials, rated their methodological quality and extracted relevant data. Although meta-analyses of primary and secondary endpoints were planned there was a high level of study heterogeneity and only one common outcome measure (SF-36) could be statistically synthesised. In addition, we conducted a narrative analysis of interventions, particularly in terms of inspecting and identifying intervention components, grouping or categorising interventions and examining potential common links and outcomes.

Main results: Twelve RCTs (comprising 1669 participants) met the eligibility criteria. We judged five studies to have a moderate risk of bias and assessed the remaining seven as having a high risk of bias. It was possible to include SF-36 physical health component scores from five studies in a meta-analysis. Participating in a MDRP was associated with an increase in SF-36 physical health component scores (mean difference (MD) 2.22, 95% confidence interval (CI) 0.12 to 4.31, P = 0.04). The findings from the narrative analysis suggested that MDRPs with a single domain or outcome focus appeared to be more successful than programmes with multiple aims. In addition, programmes that comprised participants with different types of cancer compared to cancer site-specific programmes were more likely to show positive improvements in physical outcomes. The most effective mode of service delivery appeared to be face-to-face contact supplemented with at least one follow-up telephone call. There was no evidence to indicate that MDRPs which lasted longer than six months improved outcomes beyond the level attained at six months. In addition, there was no evidence to suggest that services were more effective if they were delivered by a particular type of health professional.

Authors' conclusions: There is some evidence to support the effectiveness of brief, focused MDRPs for cancer survivors. Rigorous and methodologically sound clinical trials that include an economic analysis are required.

Multidimensional Intrusion Detection System for IEC 61850 based SCADA Networks

Emerging cybersecurity vulnerabilities in supervisory control and data acquisition (SCADA) systems are becoming urgent engineering issues for modern substations. This paper proposes a novel intrusion detection system (IDS) tailored for cybersecurity of IEC 61850 based substations. The proposed IDS integrates physical knowledge, protocol specifications and logical behaviours to provide a comprehensive and effective solution that is able to mitigate various cyberattacks. The proposed approach comprises access control detection, protocol whitelisting, model-based detection, and multi-parameter based detection. This SCADA-specific IDS is implemented and validated using a comprehensive and realistic cyber-physical test-bed and data from a real 500kV smart substation.