A method for automatically eliciting node weights in a hierarchical knowledge-based structure for reasoning with uncertainty

Hierarchical knowledge structures are frequently used within clinical decision support systems as part of the model for generating intelligent advice. The nodes in the hierarchy inevitably have varying influence on the decisionmaking processes, which needs to be reflected by parameters. If the model has been elicited from human experts, it is not feasible to ask them to estimate the parameters because there will be so many in even moderately-sized structures. This paper describes how the parameters could be obtained from data instead, using only a small number of cases. The original method [1] is applied to a particular web-based clinical decision support system called GRiST, which uses its hierarchical knowledge to quantify the risks associated with mental-health problems. The knowledge was elicited from multidisciplinary mental-health practitioners but the tree has several thousand nodes, all requiring an estimation of their relative influence on the assessment process. The method described in the paper shows how they can be obtained from about 200 cases instead. It greatly reduces the experts’ elicitation tasks and has the potential for being generalised to similar knowledge-engineering domains where relative weightings of node siblings are part of the parameter space.

An investigation into inductive parameter learning in complex hierarchical knowledge structures representing clinical expertise

This dissertation investigates the very important and current problem of modelling human expertise. This is an apparent issue in any computer system emulating human decision making. It is prominent in Clinical Decision Support Systems (CDSS) due to the complexity of the induction process and the vast number of parameters in most cases. Other issues such as human error and missing or incomplete data present further challenges. In this thesis, the Galatean Risk Screening Tool (GRiST) is used as an example of modelling clinical expertise and parameter elicitation. The tool is a mental health clinical record management system with a top layer of decision support capabilities. It is currently being deployed by several NHS mental health trusts across the UK. The aim of the research is to investigate the problem of parameter elicitation by inducing them from real clinical data rather than from the human experts who provided the decision model. The induced parameters provide an insight into both the data relationships and how experts make decisions themselves. The outcomes help further understand human decision making and, in particular, help GRiST provide more accurate emulations of risk judgements. Although the algorithms and methods presented in this dissertation are applied to GRiST, they can be adopted for other human knowledge engineering domains.

Statnote 30 : The one-way analysis of variance (ANOVA):fixed effects model

In Statnote 9, we described a one-way analysis of variance (ANOVA) ‘random effects’ model in which the objective was to estimate the degree of variation of a particular measurement and to compare different sources of variation in space and time. The illustrative scenario involved the role of computer keyboards in a University communal computer laboratory as a possible source of microbial contamination of the hands. The study estimated the aerobic colony count of ten selected keyboards with samples taken from two keys per keyboard determined at 9am and 5pm. This type of design is often referred to as a ‘nested’ or ‘hierarchical’ design and the ANOVA estimated the degree of variation: (1) between keyboards, (2) between keys within a keyboard, and (3) between sample times within a key. An alternative to this design is a 'fixed effects' model in which the objective is not to measure sources of variation per se but to estimate differences between specific groups or treatments, which are regarded as 'fixed' or discrete effects. This statnote describes two scenarios utilizing this type of analysis: (1) measuring the degree of bacterial contamination on 2p coins collected from three types of business property, viz., a butcher’s shop, a sandwich shop, and a newsagent and (2) the effectiveness of drugs in the treatment of a fungal eye infection.

Biomedical events extraction using the hidden vector state model

Objective: Biomedical events extraction concerns about events describing changes on the state of bio-molecules from literature. Comparing to the protein-protein interactions (PPIs) extraction task which often only involves the extraction of binary relations between two proteins, biomedical events extraction is much harder since it needs to deal with complex events consisting of embedded or hierarchical relations among proteins, events, and their textual triggers. In this paper, we propose an information extraction system based on the hidden vector state (HVS) model, called HVS-BioEvent, for biomedical events extraction, and investigate its capability in extracting complex events. Methods and material: HVS has been previously employed for extracting PPIs. In HVS-BioEvent, we propose an automated way to generate abstract annotations for HVS training and further propose novel machine learning approaches for event trigger words identification, and for biomedical events extraction from the HVS parse results. Results: Our proposed system achieves an F-score of 49.57% on the corpus used in the BioNLP'09 shared task, which is only 2.38% lower than the best performing system by UTurku in the BioNLP'09 shared task. Nevertheless, HVS-BioEvent outperforms UTurku's system on complex events extraction with 36.57% vs. 30.52% being achieved for extracting regulation events, and 40.61% vs. 38.99% for negative regulation events. Conclusions: The results suggest that the HVS model with the hierarchical hidden state structure is indeed more suitable for complex event extraction since it could naturally model embedded structural context in sentences.

Extracting protein-protein interactions from MEDLINE using the hidden vector state model

A major challenge in text mining for biomedicine is automatically extracting protein-protein interactions from the vast amount of biomedical literature. We have constructed an information extraction system based on the Hidden Vector State (HVS) model for protein-protein interactions. The HVS model can be trained using only lightly annotated data whilst simultaneously retaining sufficient ability to capture the hierarchical structure. When applied in extracting protein-protein interactions, we found that it performed better than other established statistical methods and achieved 61.5% in F-score with balanced recall and precision values. Moreover, the statistical nature of the pure data-driven HVS model makes it intrinsically robust and it can be easily adapted to other domains.

Estimating the SCAN*PRO model of store sales: HB, FM or just OLS?

In this paper we investigate whether consideration of store-level heterogeneity in marketing mix effects improves the accuracy of the marketing mix elasticities, fit, and forecasting accuracy of the widely-applied SCAN*PRO model of store sales. Models with continuous and discrete representations of heterogeneity, estimated using hierarchical Bayes (HB) and finite mixture (FM) techniques, respectively, are empirically compared to the original model, which does not account for store-level heterogeneity in marketing mix effects, and is estimated using ordinary least squares (OLS). The empirical comparisons are conducted in two contexts: Dutch store-level scanner data for the shampoo product category, and an extensive simulation experiment. The simulation investigates how between- and within-segment variance in marketing mix effects, error variance, the number of weeks of data, and the number of stores impact the accuracy of marketing mix elasticities, model fit, and forecasting accuracy. Contrary to expectations, accommodating store-level heterogeneity does not improve the accuracy of marketing mix elasticities relative to the homogeneous SCAN*PRO model, suggesting that little may be lost by employing the original homogeneous SCAN*PRO model estimated using ordinary least squares. Improvements in fit and forecasting accuracy are also fairly modest. We pursue an explanation for this result since research in other contexts has shown clear advantages from assuming some type of heterogeneity in market response models. In an Afterthought section, we comment on the controversial nature of our result, distinguishing factors inherent to household-level data and associated models vs. general store-level data and associated models vs. the unique SCAN*PRO model specification.

A multi-level investigation of psychological contract breach and organizational identification through the lens of perceived organizational membership:testing a moderated-mediated model

Drawing on the perceived organizational membership theoretical framework and the social identity view of dissonance theory, I examined in this study the dynamics of the relationship between psychological contract breach and organizational identification. I included group-level transformational and transactional leadership as well as procedural justice in the hypothesized model as key antecedents for organizational membership processes. I further explored the mediating role of psychological contract breach in the relationship between leadership, procedural justice climate, and organizational identification and proposed separateness–connectedness self-schema as an important moderator of the above mediated relationship. Hierarchical linear modeling results from a sample of 864 employees from 162 work units in 10 Greek organizations indicated that employees' perception of psychological contract breach negatively affected their organizational identification. I also found psychological contract breach to mediate the impact of transformational and transactional leadership on organizational identification. Results further provided support for moderated mediation and showed that the indirect effects of transformational and transactional leadership on identification through psychological contract breach were stronger for employees with a low connectedness self-schema.

Evaluation of options for energy recovery from municipal solid waste in India using the hierarchical analytical network process

In this paper a Hierarchical Analytical Network Process (HANP) model is demonstrated for evaluating alternative technologies for generating electricity from MSW in India. The technological alternatives and evaluation criteria for the HANP study are characterised by reviewing the literature and consulting experts in the field of waste management. Technologies reviewed in the context of India include landfill, anaerobic digestion, incineration, pelletisation and gasification. To investigate the sensitivity of the result, we examine variations in expert opinions and carry out an Analytical Hierarchy Process (AHP) analysis for comparison. We find that anaerobic digestion is the preferred technology for generating electricity from MSW in India. Gasification is indicated as the preferred technology in an AHP model due to the exclusion of criteria dependencies and in an HANP analysis when placing a high priority on net output and retention time. We conclude that HANP successfully provides a structured framework for recommending which technologies to pursue in India, and the adoption of such tools is critical at a time when key investments in infrastructure are being made. Therefore the presented methodology is thought to have a wider potential for investors, policy makers, researchers and plant developers in India and elsewhere. © 2013 Elsevier Ltd. All rights reserved.

Impact of high-performance work systems on individual- and branch-level performance:test of a multilevel model of intermediate linkages

We proposed and tested a multilevel model, underpinned by empowerment theory, that examines the processes linking high-performance work systems (HPWS) and performance outcomes at the individual and organizational levels of analyses. Data were obtained from 37 branches of 2 banking institutions in Ghana. Results of hierarchical regression analysis revealed that branch-level HPWS relates to empowerment climate. Additionally, results of hierarchical linear modeling that examined the hypothesized cross-level relationships revealed 3 salient findings. First, experienced HPWS and empowerment climate partially mediate the influence of branch-level HPWS on psychological empowerment. Second, psychological empowerment partially mediates the influence of empowerment climate and experienced HPWS on service performance. Third, service orientation moderates the psychological empowerment-service performance relationship such that the relationship is stronger for those high rather than low in service orientation. Last, ordinary least squares regression results revealed that branch-level HPWS influences branch-level market performance through cross-level and individual-level influences on service performance that emerges at the branch level as aggregated service performance. © 2011 American Psychological Association.

Semisupervised learning of hierarchical latent trait models for data visualization

Recently, we have developed the hierarchical Generative Topographic Mapping (HGTM), an interactive method for visualization of large high-dimensional real-valued data sets. In this paper, we propose a more general visualization system by extending HGTM in three ways, which allows the user to visualize a wider range of data sets and better support the model development process. 1) We integrate HGTM with noise models from the exponential family of distributions. The basic building block is the Latent Trait Model (LTM). This enables us to visualize data of inherently discrete nature, e.g., collections of documents, in a hierarchical manner. 2) We give the user a choice of initializing the child plots of the current plot in either interactive, or automatic mode. In the interactive mode, the user selects "regions of interest," whereas in the automatic mode, an unsupervised minimum message length (MML)-inspired construction of a mixture of LTMs is employed. The unsupervised construction is particularly useful when high-level plots are covered with dense clusters of highly overlapping data projections, making it difficult to use the interactive mode. Such a situation often arises when visualizing large data sets. 3) We derive general formulas for magnification factors in latent trait models. Magnification factors are a useful tool to improve our understanding of the visualization plots, since they can highlight the boundaries between data clusters. We illustrate our approach on a toy example and evaluate it on three more complex real data sets. © 2005 IEEE.

Hierarchical approach to project planning:the case of a petroleum pipeline construction

Projects exposed to an uncertain environment must be adapted to deal with the effective integration of various planning elements and the optimization of project parameters. Time, cost, and quality are the prime objectives of a project that need to be optimized to fulfill the owner's goal. In an uncertain environment, there exist many other conflicting objectives that may also need to be optimized. These objectives are characterized by varying degrees of conflict. Moreover, an uncertain environment also causes several changes in the project plan throughout its life, demanding that the project plan be totally flexible. Goal programming (GP), a multiple criteria decision making technique, offers a good solution for this project planning problem. There the planning problem is considered from the owner's perspective, which leads to classifying the project up to the activity level. GP is applied separately at each level, and the formulated models are integrated through information flow. The flexibility and adaptability of the models lies in the ease of updating the model parameters at the required level through changing priorities and/or constraints and transmitting the information to other levels. The hierarchical model automatically provides integration among various element of planning. The proposed methodology is applied in this paper to plan a petroleum pipeline construction project, and its effectiveness is demonstrated.

Hierarchical nanoporous solid base catalysts for biofuels

Novel macroporous solid bases have been developed as alternative clean technologies to existing commercial homogeneous catalysts for the production of biodiesel from triglycerides; the latter suffer process disadvantages including complex separation and associated saponification and engine corrosion, and are unsuitable for continuous operation. To this end, tuneable macroporous MgAl hydrotalcites have been prepared by an alkali-free route and characterised by TGA, XRD, SEM and XPS. The macropore architecture improves diffusion of bulky triglyceride molecules to the active base sites, increasing activity. Lamellar and macroporous hydrotalcites will be compared for the transesterification of both model and plant oil feedstocks, and structure-reactivity relations identified.

Health beliefs affect the correct replacement of daily disposable contact lenses:predicting compliance with the Health Belief Model and the Theory of Planned Behaviour

Purpose: To assess the compliance of Daily Disposable Contact Lenses (DDCLs) wearers with replacing lenses at a manufacturer-recommended replacement frequency. To evaluate the ability of two different Health Behavioural Theories (HBT), The Health Belief Model (HBM) and The Theory of Planned Behaviour (TPB), in predicting compliance. Method: A multi-centre survey was conducted using a questionnaire completed anonymously by contact lens wearers during the purchase of DDCLs. Results: Three hundred and fifty-four questionnaires were returned. The survey comprised 58.5% females and 41.5% males (mean age 34. ±. 12. years). Twenty-three percent of respondents were non-compliant with manufacturer-recommended replacement frequency (re-using DDCLs at least once). The main reason for re-using DDCLs was "to save money" (35%). Predictions of compliance behaviour (past behaviour or future intentions) on the basis of the two HBT was investigated through logistic regression analysis: both TPB factors (subjective norms and perceived behavioural control) were significant (p. <. 0.01); HBM was less predictive with only the severity (past behaviour and future intentions) and perceived benefit (only for past behaviour) as significant factors (p. <. 0.05). Conclusions: Non-compliance with DDCLs replacement is widespread, affecting 1 out of 4 Italian wearers. Results from the TPB model show that the involvement of persons socially close to the wearers (subjective norms) and the improvement of the procedure of behavioural control of daily replacement (behavioural control) are of paramount importance in improving compliance. With reference to the HBM, it is important to warn DDCLs wearers of the severity of a contact-lens-related eye infection, and to underline the possibility of its prevention.