Order selection of autoregressive models

The problem of determining the order of autoregressive models by Bayesian predictive densities is addressed. A criterion employing noninformative prior densities of the model parameters is derived. Simulation results which demonstrate the good performance of the criterion are presented. Comparisons with four popular approaches verify its superiority in many cases     

Primary-care-based randomised placebo-controlled trial of antibiotic treatment in acute maxillary sinusitis

BACKGROUND: The value of antibiotics in acute rhinosinusitis is uncertain. Although maxillary sinusitis is commonly diagnosed and treated in general practice, no effectiveness studies have been done on unselected primary-care patients. We used a randomised, placebo-controlled design to test the hypothesis that there would be an improvement associated with amoxycillin treatment for acute maxillary sinusitis patients presenting to general practice. METHODS: Adult patients with suspected acute maxillary sinusitis were referred by general practitioners for radiographs of the maxillary sinus. Those with radiographic abnormalities (n = 214) were randomly assigned treatment with amoxycillin (750 mg three times daily for 7 days; n = 108) or placebo (n = 106). Clinical course was assessed after 1 week and 2 weeks, and reported relapses and complications were recorded during the following year. FINDINGS: After 2 weeks, symptoms had improved substantially or disappeared in 83% of patients in the study group and 77% of patients taking placebo. Amoxycillin did not influence the clinical course of maxillary sinusitis nor the frequency of relapses during the 1-year follow-up. Radiographs had no prognostic value, nor were they an effect modifier. Side-effects were recorded in 28% of patients given amoxycillin and in 9% of those taking placebo (p < 0.01). The occurrence of relapses was similar in both groups (21 vs 17%) during the follow-up year. INTERPRETATION: Antibiotic treatment did not improve the clinical course of acute maxillary sinusitis presenting to general practice. For these patients, an initial radiographic examination is not necessary and initial management can be limited to symptomatic treatment. Whether antibiotics are necessary in more severe cases warrants further study.     

Faecal-Indicator Budgets for Recreational Coastal Waters: A Catchment Approach

This paper explores the application of a catchment faecal-indicator budget approach to the examination of faecal-indicator sources impacting on coastal recreational water quality. This involves (i) the characterization of water quality in sewage effluent and riverine inputs, and (ii) measurement and estimation of input flows; the product of these components yields the input flux of indicator organisms from each source. The impact of sewage-treatment scenarios on budgets is also investigated.     

Global bifurcations in a simple, autocatalytic reaction

The dynamics of a chemical oscillator based on an autocatalytic reaction are investigated. A simple local stability analysis of the steady states of the system is shown to be insufficient to give a complete picture and, in this case, can be misleading. The bifurcation analysis is therefore extended to search for global bifurcations of limit cycles and homoclinic bifurcations. A recipe for locating and path-following of the homoclinic orbit in parameter space is given. The whole of the parameter plane is thus mapped out, showing all of the local and global bifurcations and the full extent of the oscillatory regions