No abstract

Data currently generated in the field of nutrition is becoming increasingly complex and high-dimensional, bringing with it new methods of data analysis. The characteristics of machine learning make it suitable for such analysis and thus lends itself as an alternative tool to deal with data of this nature. Machine learning has already been applied in important problem areas in nutrition such as obesity, metabolic health, and malnutrition. Despite this, experts in nutrition are often without an understanding of machine learning, which limits its application and therefore potential to solve currently open questions. Thus, the current article aims to bridge this knowledge gap by supplying nutrition researchers with a resource to facilitate the use of ML in their research. Machine learning is first explained and distinguished from existing solutions, with key examples of applications in the nutrition literature provided. Two case studies of domains in which machine learning is particularly applicable, precision nutrition and metabolomics, are then presented. Finally, a framework is first outlined to guide the interested researcher in integrating machine learning into their work. By acting as a resource to which researchers can refer, we hope to support the integration of machine learning in the field of nutrition to facilitate modern research.

The origins, development, and status of the pi mechanism theory are reviewed. The paper is divided into four sections. In the first section Stiles's general ideas about 'color mechanisms' are examined, and it is concluded that foremost amongst these is a mathematical theory that specifies certain formal rules or laws that should govern a certain class of observations. In the case of pi mechanisms, the class of observations is that of two-color thresholds, and the defining laws are the two well-known displacement laws. Five other laws that two-color increment-threshold observations should obey, if the latter are governed by ideal pi mechanisms, are abstracted from Stiles's writings. In the second section literature pertinent to the testing of the seven Stilesian laws is reviewed, and it is asked whether or not the seven pi mechanisms of Stiles do in fact obey the laws. In the third section the relation of the pi mechanism concept to physiological concepts is examined, and its relation to the 'cone fundamental' is discussed; the evidence pertinent to the question: "Are any of the pi mechanisms of the single-fundamental type?" is then reviewed. The last section is devoted to the evolution of Stiles's ideas in the period after 1959 when Stiles's own investigations and those of others propelled him to reject the initial (1953) pi mechanism theory as an adequate characterization of the data of the two-color threshold.

to y ield a nsk order ' and thus a behaviorally tional Institutes of Health Training Grant GM-derived risk order would be obtained. Risk-01231-13. This paper has benefited from discussions taking behavior and many questions of theowith David Krantz.retical interest pertaining to the nature of Requests for reprints should be sent to Clyde H. 1 1 -1 -1 i j j Coombs, 580 Union Drive, Ann Arbor, Michigan psychological risk are then opened to study 48109.in a well-controlled way.