Mixed-effect Bayesian network reveals personal effects of nutrition

Nutrition experts know by their experience that people can react very differently to the same nutrition. If we could systematically quantify these differences, it would enable more personal dietary understanding and guidance. This work proposes a mixed-effect Bayesian network as a method for modeling the multivariate system of nutrition effects. Estimation of this network reveals a system of both population-wide and personal correlations between nutrients and their biological responses. Fully Bayesian estimation in the method allows managing the uncertainty in parameters and incorporating the existing nutritional knowledge into the model. The method is evaluated by modeling data from a dietary intervention study, called Sysdimet, which contains personal observations from food records and the corresponding fasting concentrations of blood cholesterol, glucose, and insulin. The model’s usefulness in nutritional guidance is evaluated by predicting personally if a given diet increases or decreases future levels of concentrations. The proposed method is shown to be comparable with the well-performing Extreme Gradient Boosting (XGBoost) decision tree method in classifying the directions of concentration increases and decreases. In addition to classification, we can also predict the precise concentration level and use the biologically interpretable model parameters to understand what personal effects contribute to the concentration. We found considerable personal differences in the contributing nutrients, and while these nutritional effects are previously known at a population level, recognizing their personal differences would result in more accurate estimates and more effective nutritional guidance.

Shortcomings of Transfer Entropy and Partial Transfer Entropy: Extending Them to Escape the Curse of Dimensionality

Transfer entropy (TE) captures the directed relationships between two variables. Partial transfer entropy (PTE) accounts for the presence of all confounding variables of a multivariate system and infers only about direct causality. However, the computation of partial transfer entropy involves high dimensional distributions and thus may not be robust in case of many variables. In this work, different variants of the partial transfer entropy are introduced, by building a reduced number of confounding variables based on different scenarios in terms of their interrelationships with the driving or response variable. Connectivity-based PTE variants utilizing the random forests (RF) methodology are evaluated on synthetic time series. The empirical findings indicate the superiority of the suggested variants over transfer entropy and partial transfer entropy, especially in the case of high dimensional systems. The above findings are further highlighted when applying the causality measures on financial time series.

An assessment of the algorithm difference measure in a multivariate system for production plans scheduling

In the process of designing a production plan, one of the important steps is scheduling the execution of technological operations. The schedule can be created either manually or by using software. If the schedule is compiled by software, then several schedule generation algorithms are used to eliminate possible errors. A set of such algorithms is called a "batch". It is advisable that only different algorithms should be included in the batch. This is necessary to eliminate errors of the same type. Therefore, the search for clones of algorithms in the batch is an urgent production task. To solve it a diversity metric of algorithms was developed in the course of this work. Such a metric numerically (as a percentage) determines how much the algorithms differ. This metric is based on the properties of the algorithm execution. Algorithm traces are constructed in the N-dimensional space using the obtained points. The coordinates of the trace points are the values with which the algorithm works at each step of its execution or each of the control points of the algorithm execution. An experiment was performed to confirm the correctness of this metric. Within this experiment, the trace properties of three sorting algorithms were calculated. Based on the properties obtained, indicators were determined for comparing algorithms in the metric space. The experiment confirmed the effectiveness of using the diversity metric to find clones in the algorithms batch. The scope of this metric is not limited to clone searches. It can be used as an independent indicator of software quality.

Online performance monitoring and diagnosis of multivariate systems

Online performance monitoring can be used to improve the performance of control systems in industry. The purpose of this article is to detect a performance deterioration and determine its cause in a system. In this article, two indices are used for online performance monitoring of a nonlinear multivariate system with optimally tuned proportional integral controllers. The first index is defined based on a squared distance measurement between the closed-loop system outputs and chosen set-points. The second index is a statistical index that uses all the information in the covariance matrices of the closed-loop system output data. Both indices are used and compared for performance monitoring of a quadruple-tank system. Moreover, hypothesis testing method has been used to determine the cause of the performance deterioration, so that appropriate solutions according to the cause can be applied to the system to improve the performance.

Physiological and Genetic Variations of Amorphophallus variabilis in Bojonegoro based on Glucomannan Content, Calcium Oxalate and RAPD Markers

This study had purpose to determine the glucomannan and calcium oxalate content in 7 variants of A. variabilis and their genetic relationships based on RAPD markers. Amorphophallus variabilis Tuber samples were taken from Karangdowo village, Sumberrejo Sub-district, Bojonegoro District, East Java. Each variant was analyzed for its glucomannan content by the spectrophotometric method using 3,5-DNS reagents and calcium oxalate by the 0.1N KMnO4 titration method. Leaf DNA extraction was carried out using the CTAB method. Relationship analysis used RAPD markers with 5 primers (OPA-11, OPC-04, OPU-06, OPC-07, and OPN-1). The obtained data were analyzed using the Numerical Taxonomy and Multivariate System (NTSYS-pc) version 2.1. The grouping of glucomannan content or oxalium[O1] oxalate used hierarchical clustering analysis (SPSS 16.0). This research found that the calcium oxalate content in seven variants of A. variabilis ranged from 0.01 to 0.03 g, where the variant with the lowest calcium oxalate is V6 with a value of 0.01 g and the highest is the V7 variant with a value of 0.03 g, while the glucomannan content ranges from 9 - 38%. The highest glucomannan content is V3, while the lowest is V6. Phenograms formed based on RAPD markers showed the formation of two groups of A. variabilis. Group one has two subgroups. Subgroup one consists of variants V1 and V4, while subgroup two consists of V6 and V7. Meanwhile, the second group consists of variants V2, V3, and V5. The seven variant grouping pattern of A. variabilis based on RAPD markers has no similarity to the grouping pattern based on the results of glucomannan or calcium oxalate analysis.

A New Measure of Ensemble Central Tendency

Ensemble prediction is a widely used tool in weather forecasting. In particular, the arithmetic mean (AM) of ensemble members is used to filter out unpredictable features from a forecast. AM is a pointwise statistical concept, providing the best sample-based estimate of the expected value of any single variable. The atmosphere, however, is a multivariate system with spatially coherent features characterized with strong correlations. Disregarding such correlations, the AM of an ensemble of forecasts removes not only unpredictable noise but also flattens features whose presence is still predictable, albeit with somewhat uncertain location. As a consequence, AM destroys the structure, and reduces the amplitude and variability associated with partially predictable features. Here we explore the use of an alternative concept of central tendency for the estimation of the expected feature (instead of single values) in atmospheric systems. Features that are coherent across ensemble members are first collocated to their mean position, before the AM of the aligned members is taken. Unlike earlier definitions based on complex variational minimization (field coalescence of Ravela and generalized ensemble mean of Purser), the proposed feature-oriented mean (FM) uses simple and computationally efficient vector operations. Though FM is still not a dynamically realizable state, a preliminary evaluation of ensemble geopotential height forecasts indicates that it retains more variance than AM, without a noticeable drop in skill. Beyond ensemble forecasting, possible future applications include a wide array of climate studies where the collocation of larger-scale features of interest may yield enhanced compositing results.

Bundled Causal History Interaction

Complex systems arise as a result of the nonlinear interactions between components. In particular, the evolutionary dynamics of a multivariate system encodes the ways in which different variables interact with each other individually or in groups. One fundamental question that remains unanswered is: How do two non-overlapping multivariate subsets of variables interact to causally determine the outcome of a specific variable? Here, we provide an information-based approach to address this problem. We delineate the temporal interactions between the bundles in a probabilistic graphical model. The strength of the interactions, captured by partial information decomposition, then exposes complex behavior of dependencies and memory within the system. The proposed approach successfully illustrated complex dependence between cations and anions as determinants of pH in an observed stream chemistry system. In the studied catchment, the dynamics of pH is a result of both cations and anions through mainly synergistic effects of the two and their individual influences as well. This example demonstrates the potentially broad applicability of the approach, establishing the foundation to study the interaction between groups of variables in a range of complex systems.