Learning Bayesian Networks with Bounded Tree-Width via Guided Search

Bounding the tree-width of a Bayesian network can reduce the chance of overfitting, and allows exact inference to be performed efficiently. Several existing algorithms tackle the problem of learning bounded tree-width Bayesian networks by learning from k-trees as super-structures, but they do not scale to large domains and/or large tree-width. We propose a guided search algorithm to find k-trees with maximum Informative scores, which is a measure of quality for the k-tree in yielding good Bayesian networks. The algorithm achieves close to optimal performance compared to exact solutions in small domains, and can discover better networks than existing approximate methods can in large domains. It also provides an optimal elimination order of variables that guarantees small complexity for later runs of exact inference. Comparisons with well-known approaches in terms of learning and inference accuracy illustrate its capabilities.

Willingness to pay for personalised nutrition across Europe

BACKGROUND: Personalised nutrition (PN) may promote public health. PN involves dietary advice based on individual characteristics of end users and can for example be based on lifestyle, blood and/or DNA profiling. Currently, PN is not refunded by most health insurance or health care plans. Improved public health is contingent on individual consumers being willing to pay for the service.

METHODS: A survey with a representative sample from the general population was conducted in eight European countries (N= 8233). Participants reported their willingness to pay (WTP) for PN based on lifestyle information, lifestyle and blood information, and lifestyle and DNA information. WTP was elicited by contingent valuation with the price of a standard, non-PN advice used as reference.

RESULTS: About 30% of participants reported being willing to pay more for PN than for non-PN advice. They were on average prepared to pay about 150% of the reference price of a standard, non-personalised advice, with some differences related to socio-demographic factors.

CONCLUSION: There is a potential market for PN compared to non-PN advice, particularly among men on higher incomes. These findings raise questions to what extent personalized nutrition can be left to the market or should be incorporated into public health programs.

Promoting healthy dietary behaviour through personalised nutrition: technology push or technology pull?

The notion of educating the public through generic healthy eating messages has pervaded dietary health promotion efforts over the years and continues to do so through various media, despite little evidence for any enduring impact upon eating behaviour. There is growing evidence, however, that tailored interventions such as those that could be delivered online can be effective in bringing about healthy dietary behaviour change. The present paper brings together evidence from qualitative and quantitative studies that have considered the public perspective of genomics, nutrigenomics and personalised nutrition, including those conducted as part of the EU-funded Food4Me project. Such studies have consistently indicated that although the public hold positive views about nutrigenomics and personalised nutrition, they have reservations about the service providers' ability to ensure the secure handling of health data. Technological innovation has driven the concept of personalised nutrition forward and now a further technological leap is required to ensure the privacy of online service delivery systems and to protect data gathered in the process of designing personalised nutrition therapies.

Psychological determinants of consumer acceptance of personalised nutrition in 9 European countries

OBJECTIVE: To develop a model of the psychological factors which predict people's intention to adopt personalised nutrition. Potential determinants of adoption included perceived risk and benefit, perceived self-efficacy, internal locus of control and health commitment.

METHODS: A questionnaire, developed from exploratory study data and the existing theoretical literature, and including validated psychological scales was administered to N=9381 participants from 9 European countries (Germany, Greece, Ireland, Poland, Portugal, Spain, the Netherlands, the UK, and Norway).

RESULTS: Structural equation modelling indicated that the greater participants' perceived benefits to be associated with personalised nutrition, the more positive their attitudes were towards personalised nutrition, and the greater their intention to adopt it. Higher levels of nutrition self-efficacy were related to more positive attitudes towards, and a greater expressed intention to adopt, personalised nutrition. Other constructs positively impacting attitudes towards personalised nutrition included more positive perceptions of the efficacy of regulatory control to protect consumers (e.g. in relation to personal data protection), higher self-reported internal health locus of control, and health commitment. Although higher perceived risk had a negative relationship with attitude and an inverse relationship with perceived benefit, its effects on attitude and intention to adopt personalised nutrition was less influential than perceived benefit. The model was stable across the different European countries, suggesting that psychological factors determining adoption of personalised nutrition have generic applicability across different European countries.

CONCLUSION: The results suggest that transparent provision of information about potential benefits, and protection of consumers' personal data is important for adoption, delivery of public health benefits, and commercialisation of personalised nutrition.