Predicting diffusion of innovations with self-organisation and machine learning

The main subject of this master's thesis was predicting diffusion of innovations. The prediction was done in a special case: product has been available in some countries, and based on its diffusion in those countries the prediction is done for other countries. The prediction was based on finding similar countries with Self-Organizing Map~(SOM), using parameters of countries. Parameters included various economical and social key figures. SOM was optimised for different products using two different methods: (a) by adding diffusion information of products to the country parameters, and (b) by weighting the country parameters based on their importance for the diffusion of different products. A novel method using Differential Evolution (DE) was developed to solve the latter, highly non-linear optimisation problem. Results were fairly good. The prediction method seems to be on a solid theoretical foundation. The results based on country data were good. Instead, optimisation for different products did not generally offer clear benefit, but in some cases the improvement was clearly noticeable. The weights found for the parameters of the countries with the developed SOM optimisation method were interesting, and most of them could be explained by properties of the products.

Self-Organising Maps and Correlation Analysis as a Tool to Explore Patterns in Excitation-Emission Matrix Data Sets and to Discriminate Dissolved Organic Matter Fluorescence Components

Dissolved organic matter (DOM) is a complex mixture of organic compounds, ubiquitous in marine and freshwater systems. Fluorescence spectroscopy, by means of Excitation-Emission Matrices (EEM), has become an indispensable tool to study DOM sources, transport and fate in aquatic ecosystems. However the statistical treatment of large and heterogeneous EEM data sets still represents an important challenge for biogeochemists. Recently, Self-Organising Maps (SOM) has been proposed as a tool to explore patterns in large EEM data sets. SOM is a pattern recognition method which clusterizes and reduces the dimensionality of input EEMs without relying on any assumption about the data structure. In this paper, we show how SOM, coupled with a correlation analysis of the component planes, can be used both to explore patterns among samples, as well as to identify individual fluorescence components. We analysed a large and heterogeneous EEM data set, including samples from a river catchment collected under a range of hydrological conditions, along a 60-km downstream gradient, and under the influence of different degrees of anthropogenic impact. According to our results, chemical industry effluents appeared to have unique and distinctive spectral characteristics. On the other hand, river samples collected under flash flood conditions showed homogeneous EEM shapes. The correlation analysis of the component planes suggested the presence of four fluorescence components, consistent with DOM components previously described in the literature. A remarkable strength of this methodology was that outlier samples appeared naturally integrated in the analysis. We conclude that SOM coupled with a correlation analysis procedure is a promising tool for studying large and heterogeneous EEM data sets.

Effects of manipulating Task Complexity on self-repairs during L2 oral production

This paper analyses the effects of manipulating the cognitive complexity of L2 oral tasks on language production. It specifically focuses on self-repairs, which are taken as a measure of accuracy since they denote both attention to form and an attempt at being accurate. By means of a repeated measures de- sign, 42 lower-intermediate students were asked to perform three different tasks types (a narrative, and instruction-giving task, and a decision-making task) for which two degrees of cognitive complexity were established. The narrative task was manipulated along +/− Here-and-Now, an instruction-giving task ma- nipulated along +/− elements, and the decision-making task which is manipu- lated along +/− reasoning demands. Repeated measures ANOVAs are used for the calculation of differences between degrees of complexity and among task types. One-way ANOVA are used to detect potential differences between low- proficiency and high-proficiency participants. Results show an overall effect of Task Complexity on self-repairs behavior across task types, with different be- haviors existing among the three task types. No differences are found between the self-repair behavior between low and high proficiency groups. Results are discussed in the light of theories of cognition and L2 performance (Robin- son 2001a, 2001b, 2003, 2005, 2007), L1 and L2 language production models (Levelt 1989, 1993; Kormos 2000, 2006), and attention during L2 performance (Skehan 1998; Robinson, 2002).

Self-accelerating dolomite-for-calcite replacement: Self-organized dynamics of burial dolomitization and associated mineralization

A new dynamic model of dolomitization predicts a multitude of textural, paragenetic, geochemical and other properties of burial dolomites. The model is based on two postulates, (1) that the dolomitizing brine is Mg-rich but under saturated with both calcite and dolomite, and (2) that the dolomite-for-calcite replacement happens not by dissolution-precipitation as usually assumed, but by dolomite-growth-driven pressure solution of the calcite host. Crucially, the dolomite-for-calcite replacement turns out to be self-accelerating via Ca2 : the Ca2 released by each replacement increment accelerates the rate of the next, and so on. As a result, both pore-fluid Ca2 and replacement rate grow exponentially.

Emotional intelligence compensates for low IQ and boosts low emotionality individuals in a self-presentation task

The research on emotional intelligence (EI) has focused mainly on testing the incremental validity of EI with respect to general intelligence and personality; less attention has been devoted to investigating the potential interaction effects. In a self-presentation task that required participants to obtain positive evaluations from others, individuals low in IQ but high in EI performed as well as the high IQ individuals. In addition, the low emotionality individuals performed significantly higher when also high in EI. The results extend the previous findings on the compensatory effect of EI on low IQ to the domain of interpersonal effectiveness and shed light on the effective functioning of personality traits when interpreted with the interaction of EI. Overall this study suggests that the role of EI in predicting performance might have been overlooked by checking solely for main effects and illustrates new venues for understanding the contribution of EI in explaining emotion-laden performance.

A case of suicide by self-injection of adrenaline.

Adrenaline (epinephrine) auto-injectors provide life-saving pre-hospital treatment for individuals experiencing anaphylaxis in a community setting. Errors in handling adrenaline auto-injectors, particularly by children and healthcare professionals, have been reported. Reports of adrenaline overdoses are limited in the medical literature. In most of these cases, accidental adrenaline administration results from medical error. Exogenous administration of catecholamine is responsible for cardiovascular and metabolic responses, which may cause supraventricular tachycardia, ventricular dysrhythmias and myocardial ischemia. The authors present a unique autopsy case involving a 34 year-old woman who intentionally self-injected adrenaline using an adrenaline auto-injector as part of a suicide plan. Catecholamines and metanephrines were measured in peripheral and cardiac blood as well as urine and vitreous humor. Based on the results of all postmortem investigations, the cause of death was determined to be cardiac dysrhythmia and cardiac arrest following adrenaline self-injection.

Self-expandable metal stents in the treatment of acute esophageal variceal bleeding.

Acute variceal bleeding (AVB) is a life-threatening complication in patients with cirrhosis. Hemostatic therapy of AVB includes early administration of vasoactive drugs that should be combined with endoscopic therapy, preferably banding ligation. However, failure to control bleeding or early rebleed within 5 days still occurs in 15-20% of patients with AVB. In these cases, a second endoscopic therapy may be attempted (mild bleeding in a hemodynamically stable patient) or we can use a balloon tamponade as a bridge to definitive derivative treatment (i.e., a transjugular intrahepatic portosystemic shunt). Esophageal balloon tamponade provides initial control in up to 80% of AVB, but it carries a high risk of major complications, especially in cases of long duration of tamponade (>24 h) and when tubes are inserted by inexperienced staff. Preliminary reports suggest that self-expandable covered esophageal metallic stents effectively control refractory AVB (i.e., ongoing bleeding despite pharmacological and endoscopic therapy or massive bleeding precluding endoscopic therapy) with a low incidence of complications. Thus, covered self-expanding metal stents may represent an alternative to the Sengstaken-Blakemore balloon for the temporary control of bleeding in treatment failures. Further studies are required to determine the role of this new device in AVB.

Characteristics of non-responders to self-reported questionnaires in a large inflammatory bowel disease cohort study.

BACKGROUND: A major threat to the validity of longitudinal cohort studies is non-response to follow-up, which can lead to erroneous conclusions. The objective of this study was to evaluate the profile of non-responders to self-reported questionnaires in the Swiss inflammatory bowel disease (IBD) Cohort. METHODS: We used data from adult patients enrolled between November 2006 and June 2011. Responders versus non-responders were compared according to socio-demographic, clinical and psychosocial characteristics. Odds ratio for non-response to initial patient questionnaire (IPQ) compared to 1-year follow-up questionnaire (FPQ) were calculated. RESULTS: A total of 1943 patients received IPQ, in which 331 (17%) did not respond. Factors inversely associated with non-response to IPQ were age >50 and female gender (OR = 0.37; p < 0.001 respectively OR = 0.63; p = 0.003) among Crohn's disease (CD) patients, and disease duration >16 years (OR = 0.48; p = 0.025) among patients with ulcerative colitis (UC). FPQ was sent to 1586 patients who had completed the IPQ; 263 (17%) did not respond. Risk factors of non-response to FPQ were mild depression (OR = 2.17; p = 0.003) for CD, and mild anxiety (OR = 1.83; p = 0.024) for UC. Factors inversely associated with non-response to FPQ were: age >30 years, colonic only disease location, higher education and higher IBD-related quality of life for CD, and age >50 years or having a positive social support for UC. CONCLUSIONS: Characteristics of non-responders differed between UC and CD. The risk of non-response to repetitive solicitations (longitudinal versus transversal study) seemed to decrease with age. Assessing non-respondents' characteristics is important to document potential bias in longitudinal studies.

Single centre experience of the application of self navigated 3D whole heart cardiovascular magnetic resonance for the assessment of cardiac anatomy in congenital heart disease.

BACKGROUND: For free-breathing cardiovascular magnetic resonance (CMR), the self-navigation technique recently emerged, which is expected to deliver high-quality data with a high success rate. The purpose of this study was to test the hypothesis that self-navigated 3D-CMR enables the reliable assessment of cardiovascular anatomy in patients with congenital heart disease (CHD) and to define factors that affect image quality. METHODS: CHD patients ≥2 years-old and referred for CMR for initial assessment or for a follow-up study were included to undergo a free-breathing self-navigated 3D CMR at 1.5T. Performance criteria were: correct description of cardiac segmental anatomy, overall image quality, coronary artery visibility, and reproducibility of great vessels diameter measurements. Factors associated with insufficient image quality were identified using multivariate logistic regression. RESULTS: Self-navigated CMR was performed in 105 patients (55% male, 23 ± 12y). Correct segmental description was achieved in 93% and 96% for observer 1 and 2, respectively. Diagnostic quality was obtained in 90% of examinations, and it increased to 94% if contrast-enhanced. Left anterior descending, circumflex, and right coronary arteries were visualized in 93%, 87% and 98%, respectively. Younger age, higher heart rate, lower ejection fraction, and lack of contrast medium were independently associated with reduced image quality. However, a similar rate of diagnostic image quality was obtained in children and adults. CONCLUSION: In patients with CHD, self-navigated free-breathing CMR provides high-resolution 3D visualization of the heart and great vessels with excellent robustness.

Self-reported nonadherence to antiretroviral therapy as a predictor of viral failure and mortality

OBJECTIVE: To determine the effect of nonadherence to antiretroviral therapy (ART) on virologic failure and mortality in naive individuals starting ART. DESIGN: Prospective observational cohort study. METHODS: Eligible individuals enrolled in the Swiss HIV Cohort Study, started ART between 2003 and 2012, and provided adherence data on at least one biannual clinical visit. Adherence was defined as missed doses (none, one, two, or more than two) and percentage adherence (>95, 90-95, and <90) in the previous 4 weeks. Inverse probability weighting of marginal structural models was used to estimate the effect of nonadherence on viral failure (HIV-1 viral load >500 copies/ml) and mortality. RESULTS: Of 3150 individuals followed for a median 4.7 years, 480 (15.2%) experienced viral failure and 104 (3.3%) died, 1155 (36.6%) reported missing one dose, 414 (13.1%) two doses and, 333 (10.6%) more than two doses of ART. The risk of viral failure increased with each missed dose (one dose: hazard ratio [HR] 1.15, 95% confidence interval 0.79-1.67; two doses: 2.15, 1.31-3.53; more than two doses: 5.21, 2.96-9.18). The risk of death increased with more than two missed doses (HR 4.87, 2.21-10.73). Missing one to two doses of ART increased the risk of viral failure in those starting once-daily (HR 1.67, 1.11-2.50) compared with those starting twice-daily regimens (HR 0.99, 0.64-1.54, interaction P = 0.09). Consistent results were found for percentage adherence. CONCLUSION: Self-report of two or more missed doses of ART is associated with an increased risk of both viral failure and death. A simple adherence question helps identify patients at risk for negative clinical outcomes and offers opportunities for intervention.

Measuring individuals' response quality in self-administered psychological tests : an introduction to Gendre's functional method

The functional method is a new test theory using a new scoring method that assumes complexity in test structure, and thus takes into account every correlation between factors and items. The main specificity of the functional method is to model test scores by multiple regression instead of estimating them by using simplistic sums of points. In order to proceed, the functional method requires the creation of hyperspherical measurement space, in which item responses are expressed by their correlation with orthogonal factors. This method has three main qualities. First, measures are expressed in the absolute metric of correlations; therefore, items, scales and persons are expressed in the same measurement space using the same single metric. Second, factors are systematically orthogonal and without errors, which is optimal in order to predict other outcomes. Such predictions can be performed to estimate how one would answer to other tests, or even to model one's response strategy if it was perfectly coherent. Third, the functional method provides measures of individuals' response validity (i.e., control indices). Herein, we propose a standard procedure in order to identify whether test results are interpretable and to exclude invalid results caused by various response biases based on control indices.