A mathematical model of neuromuscular adaptation to resistance training and its application in a computer simulation of accommodating loads

A large corpus of data obtained by means of empirical study of neuromuscular adaptation is currently of limited use to athletes and their coaches. One of the reasons lies in the unclear direct practical utility of many individual trials. This paper introduces a mathematical model of adaptation to resistance training, which derives its elements from physiological fundamentals on the one side, and empirical findings on the other. The key element of the proposed model is what is here termed the athlete’s capability profile. This is a generalization of length and velocity dependent force production characteristics of individual muscles, to an exercise with arbitrary biomechanics. The capability profile, a two-dimensional function over the capability plane, plays the central role in the proposed model of the training-adaptation feedback loop. Together with a dynamic model of resistance the capability profile is used in the model’s predictive stage when exercise performance is simulated using a numerical approximation of differential equations of motion. Simulation results are used to infer the adaptational stimulus, which manifests itself through a fed back modification of the capability profile. It is shown how empirical evidence of exercise specificity can be formulated mathematically and integrated in this framework. A detailed description of the proposed model is followed by examples of its application—new insights into the effects of accommodating loading for powerlifting are demonstrated. This is followed by a discussion of the limitations of the proposed model and an overview of avenues for future work.

Reasoned versus reactive prediction of behaviour: a meta-analysis of the prototype willingness model

The prototype willingness model (PWM) was designed to extend expectancy-value models of health behaviour by also including a heuristic, or social reactive pathway, to better explain health-risk behaviours in adolescents and young adults. The pathway includes prototype, i.e., images of a typical person who engages in a behaviour, and willingness to engage in behaviour. The current study describes a meta-analysis of predictive research using the PWM and explores the role of the heuristic pathway and intentions in predicting behaviour. Eighty-one studies met inclusion criteria. Overall, the PWM was supported and explained 20.5% of the variance in behaviour. Willingness explained 4.9% of the variance in behaviour over and above intention, although intention tended to be more strongly related to behaviour than was willingness. The strength of the PWM relationships tended to vary according to the behaviour being tested, with alcohol consumption being the behaviour best explained. Age was also an important moderator, and, as expected, PWM behaviour was best accounted for within adolescent samples. Results were heterogeneous even after moderators were taken into consideration. This meta-analysis provides support for the PWM and may be used to inform future interventions that can be tailored for at-risk populations.

Intervention-driven predictive framework for modeling healthcare data

Assessing prognostic risk is crucial to clinical care, and critically dependent on both diagnosis and medical interventions. Current methods use this augmented information to build a single prediction rule. But this may not be expressive enough to capture differential effects of interventions on prognosis. To this end, we propose a supervised, Bayesian nonparametric framework that simultaneously discovers the latent intervention groups and builds a separate prediction rule for each intervention group. The prediction rule is learnt using diagnosis data through a Bayesian logistic regression. For inference, we develop an efficient collapsed Gibbs sampler. We demonstrate that our method outperforms baselines in predicting 30-day hospital readmission using two patient cohorts - Acute Myocardial Infarction and Pneumonia. The significance of this model is that it can be applied widely across a broad range of medical prognosis tasks. © 2014 Springer International Publishing.

Expected shannon entropy and shannon differentiation between subpopulations for neutral genes under the finite island model

Shannon entropy H and related measures are increasingly used in molecular ecology and population genetics because (1) unlike measures based on heterozygosity or allele number, these measures weigh alleles in proportion to their population fraction, thus capturing a previously-ignored aspect of allele frequency distributions that may be important in many applications; (2) these measures connect directly to the rich predictive mathematics of information theory; (3) Shannon entropy is completely additive and has an explicitly hierarchical nature; and (4) Shannon entropy-based differentiation measures obey strong monotonicity properties that heterozygosity-based measures lack. We derive simple new expressions for the expected values of the Shannon entropy of the equilibrium allele distribution at a neutral locus in a single isolated population under two models of mutation: the infinite allele model and the stepwise mutation model. Surprisingly, this complex stochastic system for each model has an entropy expressable as a simple combination of well-known mathematical functions. Moreover, entropy- and heterozygosity-based measures for each model are linked by simple relationships that are shown by simulations to be approximately valid even far from equilibrium. We also identify a bridge between the two models of mutation. We apply our approach to subdivided populations which follow the finite island model, obtaining the Shannon entropy of the equilibrium allele distributions of the subpopulations and of the total population. We also derive the expected mutual information and normalized mutual information ("Shannon differentiation") between subpopulations at equilibrium, and identify the model parameters that determine them. We apply our measures to data from the common starling (Sturnus vulgaris) in Australia. Our measures provide a test for neutrality that is robust to violations of equilibrium assumptions, as verified on real world data from starlings.

Forests of the sea: predictive habitat modelling to assess the abundance of canopy forming kelp forests on temperate reefs

 Large brown seaweeds (kelps) form forests in temperate and boreal marine systems that serve as foundations to the structure and dynamics of communities. Mapping the distributions of these species is important to understanding the ecology of coastal environments, managing marine ecosystems (e.g., spatial planning), predicting consequences of climate change and the potential for carbon production. We demonstrate how combining seafloor mapping technologies (LiDAR and multibeam bathymetry) and models of wave energy to map the distribution and relative abundance of seaweed forests of Ecklonia radiata can provide complete coverage over hundreds of square kilometers. Using generalized linear mixed models (GLMMs), we associated observations of E. radiata abundance from video transects with environmental variables. These relationships were then used to predict the distribution of E. radiata across our 756.1km2 study area off the coast of Victoria, Australia. A reserved dataset was used to test the accuracy of these predictions. We found that the abundance distribution of E. radiata is strongly associated with depth, presence of rocky reef, curvature of the reef topography, and wave exposure. In addition, the GLMM methodology allowed us to adequately account for spatial autocorrelation in our sampling methods. The predictive distribution map created from the best GLMM predicted the abundance of E. radiata with an accuracy of 72%. The combination of LiDAR and multibeam bathymetry allowed us to model and predict E. radiata abundance distribution across its entire depth range for this study area. Using methods like those presented in this study, we can map the distribution of macroalgae species, which will give insight into ecological communities, biodiversity distribution, carbon uptake, and potential sequestration.

Predictive mapping of abalone fishing grounds using remotely-sensed LiDAR and commercial catch data

Defining the geographic extent of suitable fishing grounds at a scale relevant to resource exploitation for commercial benthic species can be problematic. Bathymetric light detection and ranging (LiDAR) systems provide an opportunity to enhance ecosystem-based fisheries management strategies for coastally distributed benthic fisheries. In this study we define the spatial extent of suitable fishing grounds for the blacklip abalone (Haliotis rubra) along 200 linear kilometers of coastal waters for the first time, demonstrating the potential for integration of remotely-sensed data with commercial catch information. Variables representing seafloor structure, generated from airborne bathymetric LiDAR were combined with spatially-explicit fishing event data, to characterize the geographic footprint of the western Victorian abalone fishery, in south-east Australia. A MaxEnt modeling approach determined that bathymetry, rugosity and complexity were the three most important predictors in defining suitable fishing grounds (AUC = 0.89). Suitable fishing grounds predicted by the model showed a good relationship with catch statistics within each sub-zone of the fishery, suggesting that model outputs may be a useful surrogate for potential catch.

A predictive framework for modeling healthcare data with evolving clinical interventions

Medical interventions critically determine clinical outcomes. But prediction models either ignore interventions or dilute impact by building a single prediction rule by amalgamating interventions with other features. One rule across all interventions may not capture differential effects. Also, interventions change with time as innovations are made, requiring prediction models to evolve over time. To address these gaps, we propose a prediction framework that explicitly models interventions by extracting a set of latent intervention groups through a Hierarchical Dirichlet Process (HDP) mixture. Data are split in temporal windows and for each window, a separate distribution over the intervention groups is learnt. This ensures that the model evolves with changing interventions. The outcome is modeled as conditional, on both the latent grouping and the patients' condition, through a Bayesian logistic regression. Learning distributions for each time-window result in an over-complex model when interventions do not change in every time-window. We show that by replacing HDP with a dynamic HDP prior, a more compact set of distributions can be learnt. Experiments performed on two hospital datasets demonstrate the superiority of our framework over many existing clinical and traditional prediction frameworks.

Keeping up with innovation: a predictive framework for modeling healthcare data with evolving clinical interventions

Medical outcomes are inexorably linked to patient illness and clinical interventions. Interventions change the course of disease, crucially determining outcome. Traditional outcome prediction models build a single classifier by augmenting interventions with disease information. Interventions, however, differentially affect prognosis, thus a single prediction rule may not suffice to capture variations. Interventions also evolve over time as more advanced interventions replace older ones. To this end, we propose a Bayesian nonparametric, supervised framework that models a set of intervention groups through a mixture distribution building a separate prediction rule for each group, and allows the mixture distribution to change with time. This is achieved by using a hierarchical Dirichlet process mixture model over the interventions. The outcome is then modeled as conditional on both the latent grouping and the disease information through a Bayesian logistic regression. Experiments on synthetic and medical cohorts for 30-day readmission prediction demonstrate the superiority of the proposed model over clinical and data mining baselines.

A domain-specific approach for assessing physical activity efficacy in adolescents: from scale conception to predictive validity

Objective: To develop and test the utility of a domain-specific physical activity efficacy scale in adolescents for predicting physical activity behaviour. Design: Two independent studies were conducted. Study 1 examined the psychometric properties of a newly constructed Domain-Specific Physical Activity Efficacy Questionnaire (DSPAEQ) and study 2 tested the utility of the scale for predicting leisure- and school-time physical activity. Methods: In study 1, descriptive physical activity data were used to generate scale items. The scales factor structure and internal consistency were tested in a sample of 272 adolescents. A subsequent sample of Canadian (N = 104) and New Zealand (N = 29) adolescents, was recruited in study 2 to explore the scale's predictive validity using a subjective measure of leisure- and school-time physical activity. Results: A principle axis factor analysis in study 1 revealed a 26-item, five-factor coherent and interpretable solution; representative of leisure and recreation, household, ambulatory, transportation, and school physical activity efficacy constructs, respectively. The five-factor solution explained 81% of the response variance. In study 2 the domain-specific efficacy model explained 16% and 1% of leisure- and school-time physical activity response variance, respectively, with leisure time physical activity efficacy identified as a unique and significant contributor of leisure-time physical activity. Conclusion: Study 1 provides evidence for the tenability of a five factor DSPEAQ, while study 2 shows that the DSPEAQ has utility in predicting domain-specific physical activity. This latter finding underscores the importance of scale correspondence between the behavioural elements (leisure-time physical activity) and cognitive assessment of those elements (leisure-time physical activity efficacy).

An application of an extended effort-reward imbalance model to police absenteeism behaviour

Purpose – Frequent absences from work can be highly disruptive, whilst also potentially indicating problematic working conditions that can lead to increased withdrawal behaviour. The purpose of this paper is to test the predictive capability of an expanded effort-reward imbalance model on employee absenteeism within the context of policing.

Design/methodology/approach – Three separate reward systems are identified by the effort-reward imbalance model. In this study, the authors assessed these individual components for their contribution to officer withdrawal behaviour in the form of absenteeism frequency. Data were gathered from a sample of operational officers (n=553) within a large Australian police agency.

Findings – Findings indicate that there was a strong influence of social rewards such as social support and recognition in the workplace on officer absenteeism rates. Low workload was associated with a higher frequency of absenteeism suggesting a potential underloading effect. There were a number of significant interactions providing support for the effort-reward imbalance mechanism and the separation of the reward construct. Security rewards were particularly influential and significantly moderated the relationship between effort and absenteeism.

Research limitations/implications – Differential effects of occupational rewards were identified in the study, indicating that there are significant opportunities for expansion of the effort-reward imbalance model along with opportunities for HRM practitioners in terms of employee recognition and remuneration programmes. This research was focused on a specific sample of operational officers, therefore should be expanded to include multiple occupational groups.

Originality/value – This paper considers and expanded model of worker strain and contributes a longitudinal assessment of the association between perceived effort and reward systems and worker absenteeism.

Predictive motion planning for AUVs subject to strong time-varying currents and forecasting uncertainties

This paper presents a novel path planning method for minimizing the energy consumption of an autonomous underwater vehicle subjected to time varying ocean disturbances and forecast model uncertainty. The algorithm determines 4-Dimensional path candidates using Nonlinear Robust Model Predictive Control (NRMPC) and solutions optimised using A∗-like algorithms. Vehicle performance limits are incorporated into the algorithm with disturbances represented as spatial and temporally varying ocean currents with a bounded uncertainty in their predictions. The proposed algorithm is demonstrated through simulations using a 4-Dimensional, spatially distributed time-series predictive ocean current model. Results show the combined NRMPC and A∗ approach is capable of generating energy-efficient paths which are resistant to both dynamic disturbances and ocean model uncertainty.

In-situ evaluation of predictive models for H₂S gas emission and the performance of optimal dosage of suppressing chemicals in a laboratory-scale sewer

This in-situ analysis quantifies hydrogen sulfide gas emission from a simulated sewerage system, with varying slopes between 0.5% and 1.5%, under the dosing of certain mitigating chemicals. A portable H₂S gas detector (OdaLog) was employed to record the gaseous phase concentration of hydrogen sulfide. The investigation was comprised of three interrelated phases. In the first stage, precision of four prediction models for H₂S gas emission from a laboratory-synthesized wastewater was assessed. It was found that the model suggested by Lahav fitted the experimental results accurately. Second phase explorations included jar tests to obtain the optimal dosage of four hydrogen sulfide suppressing chemicals, being Mg(OH)₂, NaOH, Ca(NO₃)₂, and FeCl₂. In the third stage, the optimal dosage of chemicals was introduced into the experimental sewerage system, with the OdaLog continuously monitoring the H₂S gas emission. According to a baseline (experiments with no chemical addition), it was found that NaOH and Mg(OH)₂ performed very good in mitigating the release of H₂S gas, while Ca(NO₃)₂ was not effective most probably due to the absence of biological activity. Furthermore, interpretation of OdaLog data through the optimum emission prediction model revealed that higher sewer slope led to more emission.

Guidelines for developing and reporting machine learning predictive models in biomedical research : a multidisciplinary view

BACKGROUND: As more and more researchers are turning to big data for new opportunities of biomedical discoveries, machine learning models, as the backbone of big data analysis, are mentioned more often in biomedical journals. However, owing to the inherent complexity of machine learning methods, they are prone to misuse. Because of the flexibility in specifying machine learning models, the results are often insufficiently reported in research articles, hindering reliable assessment of model validity and consistent interpretation of model outputs.

OBJECTIVE: To attain a set of guidelines on the use of machine learning predictive models within clinical settings to make sure the models are correctly applied and sufficiently reported so that true discoveries can be distinguished from random coincidence.

METHODS: A multidisciplinary panel of machine learning experts, clinicians, and traditional statisticians were interviewed, using an iterative process in accordance with the Delphi method.

RESULTS: The process produced a set of guidelines that consists of (1) a list of reporting items to be included in a research article and (2) a set of practical sequential steps for developing predictive models.

CONCLUSIONS: A set of guidelines was generated to enable correct application of machine learning models and consistent reporting of model specifications and results in biomedical research. We believe that such guidelines will accelerate the adoption of big data analysis, particularly with machine learning methods, in the biomedical research community.