This review evaluates the methodological quality of current front-of-pack labeling research and discusses future research challenges. Peer-reviewed articles were identified using a computerized search of the databases PubMed andWeb of Science (ISI) from1990 to February 2011; reference lists fromkey published articleswere used as well. The quality of the 31 included studies was assessed. The results showed that the methodological quality of published front-of-pack labeling research is generally low to mediocre; objective observational data-based consumer studies were of higher quality than consumer studies relying on self-reports. Experimental studies that included a control group were lacking. The review further revealed a lack of a validated methodology to measure the use of front-of-pack labels and the effects of these labels in real-life settings. In conclusion, few methodologically sound front-of-pack labeling studies are presently available. The highest methodological quality and the greatest public health relevance are achieved by measuring the health effects of front-of-pack labels using biomarkers in a longitudinal, randomized, controlled design in a real-life setting.
LINK
Abstract: Existing frailty models have enhanced research and practice; however, none of the models accounts for the perspective of older adults upon defining and operationalizing frailty. We aim to propose a mixed conceptual model that builds on the integral model while accounting for older adults’ perceptions and lived experiences of frailty. We conducted a traditional literature review to address frailty attributes, risk factors, consequences, perceptions, and lived experiences of older adults with frailty. Frailty attributes are vulnerability/susceptibility, aging, dynamic, complex, physical, psychological, and social. Frailty perceptions and lived experience themes/subthemes are refusing frailty labeling, being labeled “by others” as compared to “self-labeling”, from the perception of being frail towards acting as being frail, positive self-image, skepticism about frailty screening, communicating the term “frail”, and negative and positive impacts and experiences of frailty. Frailty risk factors are classified into socio-demographic, biological, physical, psychological/cognitive, behavioral, and situational/environmental factors. The consequences of frailty affect the individual, the caregiver/family, the healthcare sector, and society. The mixed conceptual model of frailty consists of interacting risk factors, interacting attributes surrounded by the older adult’s perception and lived experience, and interacting consequences at multiple levels. The mixed conceptual model provides a lens to qualify frailty in addition to quantifying it.
DOCUMENT
Active learning has become an increasingly popular method for screening large amounts of data in systematic reviews and meta-analyses. The active learning process continually improves its predictions on the remaining unlabeled records, with the goal of identifying all relevant records as early as possible. However, determining the optimal point at which to stop the active learning process is a challenge. The cost of additional labeling of records by the reviewer must be balanced against the cost of erroneous exclusions. This paper introduces the SAFE procedure, a practical and conservative set of stopping heuristics that offers a clear guideline for determining when to end the active learning process in screening software like ASReview. The eclectic mix of stopping heuristics helps to minimize the risk of missing relevant papers in the screening process. The proposed stopping heuristic balances the costs of continued screening with the risk of missing relevant records, providing a practical solution for reviewers to make informed decisions on when to stop screening. Although active learning can significantly enhance the quality and efficiency of screening, this method may be more applicable to certain types of datasets and problems. Ultimately, the decision to stop the active learning process depends on careful consideration of the trade-off between the costs of additional record labeling against the potential errors of the current model for the specific dataset and context.
LINK
Active learning has become an increasingly popular method for screening large amounts of data in systematic reviews and meta-analyses. The active learning process continually improves its predictions on the remaining unlabeled records, with the goal of identifying all relevant records as early as possible. However, determining the optimal point at which to stop the active learning process is a challenge. The cost of additional labeling of records by the reviewer must be balanced against the cost of erroneous exclusions. This paper introduces the SAFE procedure, a practical and conservative set of stopping heuristics that offers a clear guideline for determining when to end the active learning process in screening software like ASReview. The eclectic mix of stopping heuristics helps to minimize the risk of missing relevant papers in the screening process. The proposed stopping heuristic balances the costs of continued screening with the risk of missing relevant records, providing a practical solution for reviewers to make informed decisions on when to stop screening. Although active learning can significantly enhance the quality and efficiency of screening, this method may be more applicable to certain types of datasets and problems. Ultimately, the decision to stop the active learning process depends on careful consideration of the trade-off between the costs of additional record labeling against the potential errors of the current model for the specific dataset and context.
MULTIFILE
De Europese Unie heeft zich ten doel gesteld om de gevolgen van klimaatverandering in te perken. Als gevolg daarvan zal de duurzame energieproductie in de komende jaren (naar verwachting) toenemen en de productie van fossiele energie afnemen. De verwachting is dat een groot gedeelte van deze duurzame energieproductie uit intermitterende energiebronnen zal bestaan zoals wind- en zonne-energie, al blijft daarbij het probleem dat energie uit dergelijke bronnen niet altijd geleverd kan worden op het moment dat er vraag naar is. De energiemarkt heeft behoefte aan flexibiliteit en energieopslag kan daarin voorzien. Opslagtechnologieën bieden de mogelijkheid om overproductie van intermitterende bronnen op te slaan en daarmee de vraag naar energie op te vangen op momenten van onderproductie. Om te bepalen welke opslagtechnologie het meest geschikt is voor welke situatie, heeft de Hanzehogeschool Groningen in opdracht van Netbeheer Nederland een opslaglabel ontwikkeld dat dit mogelijk maakt.
DOCUMENT
About 20% of energy intake in the Netherlands is consumed out-of-home. Eating out-of-home is associated with higher energy intake and poorer nutrition. Menu labeling can be considered a promising instrument to improve dietary choices in the out-of-home sector. Effectiveness depends on the presentation format of the label and its attractiveness and usability to restaurant guests and restaurant owners. This exploratory study investigated which menu labeling format would be mostly appreciated by (a) (potential) restaurant guests (n386) and (b) the uninvestigated group of restaurant owners (n41) if menu labeling would be implemented in Dutch full-service restaurants. A cross-sectional survey design was used to investigate three distinct menu labeling formats: a simple health logo; (star) ranking and calorie information. Questionnaires were used as study tool. Ranking has been shown to be the most appreciated menu labeling format by both (potential) restaurant guests and owners. Statistical analysis showed that label preference of potential restaurant guests was significantly associated with age, possibly associated with level of education, and not associated with health consciousness. In summary, we found that ranking is the most appreciated menu label format according to both (potential) restaurant guests and restaurant owners, suggesting it to be a promising way to improve healthy eating out-of-home.
LINK
Human activity recognition system is of great importance in robot-care scenarios. Typically, training such a system requires activity labels to be both completely and accurately annotated. In this paper, we go beyond such restriction and propose a learning method that allow labels to be incomplete and uncertain. We introduce the idea of soft labels which allows annotators to assign multiple, and weighted labels to data segments. This is very useful in many situations, e.g., when the labels are uncertain, when part of the labels are missing, or when multiple annotators assign inconsistent labels. We formulate the activity recognition task as a sequential labeling problem. Latent variables are embedded in the model in order to exploit sub-level semantics for better estimation. We propose a max-margin framework which incorporate soft labels for learning the model parameters. The model is evaluated on two challenging datasets. To simulate the uncertainty in data annotation, we randomly change the labels for transition segments. The results show significant improvement over the state-of-the-art approach.
DOCUMENT
The Nutri-Score front-of-pack label, which classifies the nutritional quality of products in one of 5 classes (A to E), is one of the main candidates for standardized front-of-pack labeling in the EU. The algorithm underpinning the Nutri-Score label is derived from the Food Standard Agency (FSA) nutrient profile model, originally a binary model developed to regulate the marketing of foods to children in the UK. This review describes the development and validation process of the Nutri-Score algorithm. While the Nutri-Score label is one of the most studied front-of-pack labels in the EU, its validity and applicability in the European context is still undetermined. For several European countries, content validity (i.e., ability to rank foods according to healthfulness) has been evaluated. Studies showed Nutri-Score's ability to classify foods across the board of the total food supply, but did not show the actual healthfulness of products within different classes. Convergent validity (i.e., ability to categorize products in a similar way as other systems such as dietary guidelines) was assessed with the French dietary guidelines; further adaptations of the Nutri-Score algorithm seem needed to ensure alignment with food-based dietary guidelines across the EU. Predictive validity (i.e., ability to predict disease risk when applied to population dietary data) could be re-assessed after adaptations are made to the algorithm. Currently, seven countries have implemented or aim to implement Nutri-Score. These countries appointed an international scientific committee to evaluate Nutri-Score, its underlying algorithm and its applicability in a European context. With this review, we hope to contribute to the scientific and political discussions with respect to nutrition labeling in the EU.
DOCUMENT
We present a novel hierarchical model for human activity recognition. In contrast with approaches that successively recognize actions and activities, our approach jointly models actions and activities in a unified framework, and their labels are simultaneously predicted. The model is embedded with a latent layer that is able to capture a richer class of contextual information in both state-state and observation-state pairs. Although loops are present in the model, the model has an overall linear-chain structure, where the exact inference is tractable. Therefore, the model is very efficient in both inference and learning. The parameters of the graphical model are learned with a structured support vector machine. A data-driven approach is used to initialize the latent variables; therefore, no manual labeling for the latent states is required. The experimental results from using two benchmark datasets show that our model outperforms the state-of-the-art approach, and our model is computationally more efficient.
DOCUMENT
