Nuisance Factors Research Articles

Commercial beef farms excelling in terminal and maternal genetic merit generate more gross profit.

Validation of beef total merit breeding indexes for improving performance and profitability has previously been undertaken at the individual animal level; however, no herd-level validation of beef genetic merit and profit has been previously investigated. The objective of the present study was to quantify the relationship between herd profitability and both herd-average terminal and maternal genetic merit across 1,311 commercial Irish beef herds. Herd-level physical and financial performance data were available from a financial benchmarking tool used by Irish farmers and their extension advisors. Animal genetic merit data originated from the Irish Cattle Breeding Federation who undertake the national beef and dairy genetic evaluations. Herd-average genetic merit variables included the terminal index of young animals, the maternal index of dams, and the terminal index of service sires. The herds represented three production systems: 1) cow-calf to beef, 2) cow-calf to weanling/yearling, and 3) weanling/yearling to beef. Associations between herd financial performance metrics and herd average genetic merit variables were quantified using a series of linear mixed models with year, production system, herd size, stocking rate, concentrate input, and the two-way interactions between production system and herd size, stocking rate, and concentrate input included as nuisance factors. Herd nested within the county of Ireland (n = 26) was included as a repeated effect. Herds with young cattle excelling in terminal index enjoyed greater gross and net profit per hectare (ha), per livestock unit (LU), and per kg net live-weight output. The change in gross profit per LU per unit change in the terminal index of young animals was €1.41 (SE = 0.23), while the respective regression coefficient for net profit per LU was €1.37 (SE = 0.30); the standard deviation of the terminal index is €37. Herd-average dam maternal index and sire terminal index were both independently positively associated with gross profit per ha and gross profit per LU. Each one unit increase in dam maternal index (standard deviation of €38) was associated with a €1.40 (SE = 0.48) and €0.76 (SE = 0.29) greater gross profit per ha and per LU, respectively. Results from the present study at the herd-level concur with previous validation studies at the individual animal level thus instilling further confidence among stakeholders as to the expected improvement in herd profitability with improving genetic merit.

Estimating fuel-efficiency while accounting for dynamic fleet management: Testing the effects of fuel price signals and fleet rollover

Fuel efficiency has become one of the most important policy goals for airline operations management. This study develops an econometric model to estimate airlines’ fuel burn determinants, aiming to test the hypothesis of fuel price signals on carriers’ incentives for energy-saving efforts. We propose a novel high-dimensional sparse IV-LASSO method to account for airline dynamic fleet management. Our model also provides controls for thousands of nuisance factors related to route direction, airway congestion, and aircraft model specificities, allowing for flexible time-varying unobserved effects of flight stages. The results show that energy intensity reduction induced by price increases was observed 3–4 years earlier, possibly due to fleet rollover and fleet modernization. Furthermore, although new competition may create incentives for improved fuel management to control costs, our results suggest that the overall effect of market deconcentration is unclear and may produce the unintended consequence of worse performance, as measured in some key fuel efficiency dimensions.

Learning Invariant Representation of Tasks for Robust Surgical State Estimation

Surgical state estimators in robot-assisted surgery (RAS)-especially those trained via learning techniques-rely heavily on datasets that capture surgeon actions in laboratory or real-world surgical tasks. Real-world RAS datasets are costly to acquire, are obtained from multiple surgeons who may use different surgical strategies, and are recorded under uncontrolled conditions in highly complex environments. The combination of high diversity and limited data calls for new learning methods that are robust and invariant to operating conditions and surgical techniques. We propose StiseNet, a Surgical Task Invariance State Estimation Network with an invariance induction framework that minimizes the effects of variations in surgical technique and operating environments inherent to RAS datasets. StiseNet's adversarial architecture learns to separate nuisance factors from information needed for surgical state estimation. StiseNet is shown to outperform state-of-the-art state estimation methods on three datasets (including a new real-world RAS dataset: HERNIA-20).

A commentary on establishing norms for error-related brain activity during the arrow flanker task among young adults

We suggest that a large data set for the error-related negativity (ERN) and error positivity (Pe) components of the scalp-recorded event-related brain potential (ERP) recently published as normative is not ready for such use in research and, especially, clinical application. Such efforts are challenged by an incomplete understanding of the functional significance of between-person differences in amplitudes and of nuisance factors that contribute to amplitude differences, a lack of standardization of methods, and the use of a convenience sample for the potentially normative database. To move ERPs toward standardization and useful norms, we encourage more research on the meaning of differences in ERN scores, including factors that influence between- and within-person variation, and the dissemination of protocols for data collection and processing.

Short communication: Differences in genetic merit for visually-assessed body condition score materialises as phenotypic differences in tactile-based body condition score in commercial dairy cows

Body condition score (BCS) is a known risk factor for cow health and well-being. Many different BCS scales and systems for assessment exist;while the scales used for assessing BCS vary, differences in how BCS is assessed (i.e., visual versus visual plus tactile) and the extent of training and experience of the assessor (i.e., professional linear classifiers versus producers) also contributes to the underlying variability. Registered dairy cows globally are routinely assessed for linear type traits which describe biological extremes in the morphological attributes; BCS and a correlated trait angularity are within this suite of traits assessed. These linear-type data are used to generate estimates of genetic merit (predicted transmitting ability), but how these estimates manifest themselves as phenotypic differences when assessed by producers on commercial multiparous cows has never been quantified. To evaluate this, 58440 phenotypic BCS records from 48823 lactations in 38608 cows were used. Associations were undertaken using linear mixed models relating phenotypic BCS to genetic merit after accounting for nuisance factors. Differences in genetic merit for either BCS or angularity (assessed visually by professionals on a 1 to 9 scale just once during lactation in primiparous registered cows) translated to phenotypic difference in BCS (assessed by producers using both tactile and visual assessment on a 1 to 5 scale across lactation in commercial dairy cows). The partial correlation between test phenotypic BCS and genetic merit for either BCS or angularity was 0.13 and 0.10, respectively. Based on the model coefficients estimated in the present study, the mean expected difference in phenotypic BCS on a 1 to 5 scale between the top and bottom 10% on genetic merit for BCS or angularity was 0.28 and 0.31 units, respectively. Results from the present study clearly provide confidence that genetic merit for BCS or angularity based on a single visual assessment in primiparous cows is useful to breed for cows of better body condition, irrespective of stage of lactation or parity.

Behavioural repeatability in larval Limnephilus lunatus Curtis, 1834 (Trichoptera) in an open-field test

This article investigates inter-individual repeatability in distance moved in an open-field test for larval Limnephilus lunatus Curtis, 1834. Repeatability across four trials (two-day trial intervals) was comparable to previous studies on arthropod species (repeatability: R = 0.37), indicating that L. lunatus is a suitable model species in this research field. Two potential nuisance factors were corrected for: (1) progressively declining activity over consecutive trials and (2) case mass:body mass ratio, affecting activity negatively. These factors require consideration in behavioural experiments on larval caddisflies. Pairwise correlations of distance moved among trial days showed that behaviour in the first trial did not correspond well with behaviour in the following trials. Re-analysing the data using only trials 2 to 4 increased the repeatability (repeatability: R = 0.50), suggesting that future studies should consider not including data derived from initial trials, as the initial trial may constitute a different context than the following ones.

Mutual Information Regularized Feature-Level Frankenstein for Discriminative Recognition.

Deep learning recognition approaches can potentially perform better if we can extract a discriminative representation that controllably separates nuisance factors. In this paper, we propose a novel approach to explicitly enforce the extracted discriminative representation d, extracted latent variation l (e,g., background, unlabeled nuisance attributes), and semantic variation label vector s (e.g., labeled expressions/pose) to be independent and complementary to each other. We can cast this problem as an adversarial game in the latent space of an auto-encoder. Specifically, with the to-be-disentangled s, we propose to equip an end-to-end conditional adversarial network with the ability to decompose an input sample into d and l. However, we argue that maximizing the cross-entropy loss of semantic variation prediction from d is not sufficient to remove the impact of s from d, and that the uniform-target and entropy regularization are necessary. A collaborative mutual information regularization framework is further proposed to avoid unstable adversarial training. It is able to minimize the differentiable mutual information between the variables to enforce independence. The proposed discriminative representation inherits the desired tolerance property guided by prior knowledge of the task. Our proposed framework achieves top performance on diverse recognition tasks, including digits classification, large-scale face recognition on LFW and IJB-A datasets, and face recognition tolerant to changes in lighting, makeup, disguise, etc.

An Extended Youden Design in Biological Assays

Cell-based biological assays (also called bioassays) are an essential analytical tool used for quality assessment of biologic drug substances and drug products. In particular, they are used to determine the potency compared against a standard preparation under assumptions of similarity. In the design of bioassays, orthogonality between preparations and nuisance factors is desired. Block designs are an efficient way to assay multiple preparations simultaneously and mitigate bias from multiple sources, such as plates, locations within plates, and operators. The construction of a block design of a bioassay is challenging because of the treatment/control setup and the presence of multiple block effects. In many cases, absolute orthogonality is not feasible, and consequently, we need a design that achieves “near-orthogonality.” We propose an extended Youden design with mathematical construction such that near-orthogonality can be achieved in the presence of these multiple confounding effects. We also provide an efficient algorithm implementation with R. Numerical results show that the extended Youden design is the optimal design for the given scenario when absolute orthogonality is not achievable.

Investigating the impact of coercive exposure on web-advertisement performance: Focusing on the mediating role of advertisement fit

These days, advertisers make every effort to attract users through coercive exposure, so as to achieve web-advertisement performance. However, the validity of crediting actual performances through forced exposure is a moot point, because accompanied unwanted digressive advertisements and unintended mouse clicks imposed on the user, are nuisance factors that can culminate in a negative way. Hence, this study investigates advertisement performance in an elaborate way by focusing on varying effects of coercive exposures. Moreover, when considering advertisement performance, it is appropriate to examine advertisement fit, that is, whether the advertising contents match with services provided by the host website. The results showed that, when subjectively forced to frequent unsolicited advertisement, a user’s negative attitude to the respective advertisement greatly impacts on advertisement performance. Furthermore, if the level of advertisement fit is low, negative attitude is strengthened. This study provides guidelines on more appropriate ways for performing web advertisement, to achieve a genuine performance of advertisement and provide practical web-advertisement strategies for advertisers.

Learning Credit Assignment.

Deep learning has achieved impressive prediction accuracies in a variety of scientific and industrial domains. However, the nested nonlinear feature of deep learning makes the learning highly nontransparent, i.e., it is still unknown how the learning coordinates a huge number of parameters to achieve decision-making. To explain this hierarchical credit assignment, we propose a mean-field learning model by assuming that an ensemble of subnetworks, rather than a single network, is trained for a classification task. Surprisingly, our model reveals that apart from some deterministic synaptic weights connecting two neurons at neighboring layers, there exists a large number of connections that can be absent, and other connections can allow for a broad distribution of their weight values. Therefore, synaptic connections can be classified into three categories: very important ones, unimportant ones, and those of variability that may partially encode nuisance factors. Therefore, our model learns the credit assignment leading to the decision and predicts an ensemble of sub-networks that can accomplish the same task, thereby providing insights toward understanding the macroscopic behavior of deep learning through the lens of distinct roles of synaptic weights.

A Search for Suitable Growing Environment for Sesame Production in Nigeria

The economic relevance of farm management practices to the production of sesame crops in Nigeria includes preventing and combatting food crises, actualization and realization of national food security as well as enhancing gross domestic product. This research seeks to examine the edaphology of some soil types in relation to the productivity of sesame crops in Nigeria. The experiment was conducted in statistical farm of Hussaini Adamu Federal Polytechnic, Kazaure, Jigawa State, Nigeria during 2019 growing season. The experimental design was Completely Randomized Design with one hundred replications. There were three types of soils investigated; Clay, Sandy, and Loamy were compared with one another. The soils were made free of any nuisance factors effects. The chlorophyll content of the plants was read using Konica Minolta chlorophyll meter SPAD-502 plus, and the data analyzed using one-way ANOVA and the Statistical Package for Social Sciences (SPSS) version 20. The result of the analysis indicates a higher significant effect of Sandy soil (M = 11.20, SD = 2.37) on the yield of sesame crops in this region as compared to the effects of the Clay (M = 3.60, SD = 0.89), and Loamy (M = 6.86, SD = 3.39) soils. The overall ANOVA test indicates that growing environment does have effect on sesame production, F (2, 27) = 6.70, p = 0.00. However, the result indicates insignificant chlorophyll mean difference between Clay and Loamy soils. It is therefore evident from these results and the reviewed literatures that the suitable growing environment for sesame production in Nigeria is fertile Sandy soil that is deep, light textured, well-drained and that is exposed to an average temperature of 25oC to 37oC. The crop is also tolerant to draught, but not at the germination and seedling stages, water logging, and excessive rain fall while it requires 90–120 frost-free days to achieve optimal yields in cold regions. The challenges which stand in the way of increasing the productivity and quality of sesame crops need to be overcome. As with other crops, which is a major export crop in many countries, these crops should be given more research attention.

Functional engagement of white matter in resting-state brain networks

The topological characteristics of functional networks, derived from measurements of resting-state connectivity in gray matter (GM), are associated with individual cognitive abilities or specific dysfunctions. However, blood oxygen level-dependent (BOLD) signals in white matter (WM) are usually ignored or even regressed out as nuisance factors in the data analyses that underlie network models. Recent studies have demonstrated reliable detection of WM BOLD signals and imply these reflect associated neural activities. Here we evaluate quantitatively the contributions of individual WM voxels to the identification of functional networks, which we term their engagement (or conceptually, their importance). We quantify the engagement by measuring the reductions of connectivity, produced by ignoring the signal fluctuations within each WM voxel, with respect to both the entire network (global) or a single GM node (local). We observed highly reproducible spatial distributions of global engagement maps, as well as a trend toward increased relevance of deep WM voxels at delayed times. Local engagement maps exhibit homogeneous spatial distributions with respect to internal nodes that constitute a well-recognized sub-functional network, but inhomogeneous distributions with respect to other nodes. WM voxels show distinct distributions of engagement depending on their anatomical locations. These findings demonstrate the important role of WM in network modeling, thus supporting the need for changes of conventional views that WM signal variations represent only physiological noise.

Model selection for the North American Breeding Bird Survey.

The North American Breeding Bird Survey (BBS) provides data that can be used in complex, multiscale analyses of population change, while controlling for scale-specific nuisance factors. Many alternative models can be fit to the data, but most model selection procedures are not appropriate for hierarchical models. Leave-one-out cross-validation (LOOCV), in which relative model fit is assessed by omitting an observation and assessing the prediction of a model fit using the remainder of the data, provides a reasonable approach for assessing models, but is time consuming and not feasible to apply for all observations in large data sets. We report the first large-scale formal model selection for BBS data, applying LOOCV to stratified random samples of observations from BBS data. Our results are for 548 species of North American birds, comparing the fit of four alternative models that differ in year effect structures and in descriptions of extra-Poisson overdispersion. We use a hierarchical model among species to evaluate posterior probabilities that models are best for individual species. Models in which differences in year effects are conditionally independent (D models) were generally favored over models in which year effects are modeled by a slope parameter and a random year effect (S models), and models in which extra-Poisson overdispersion effects are independent and t-distributed (H models) tended to be favored over models where overdispersion was independent and normally distributed. Our conclusions lead us to recommend a change from the conventional S model to D and H models for the vast majority of species (544/548). Comparison of estimated population trends based on the favored model relative to the S model currently used for BBS summaries indicates no consistent differences in estimated trends. Of the 18 species that showed large differences in estimated trends between models, estimated trends from the default S model were more extreme, reflecting the influence of the slope parameter in that model for species that are undergoing large population changes. WAIC, a computationally simpler alternative to LOOCV, does not appear to be a reliable alternative to LOOCV.

Scientific Regulatory Policy Committee Points to Consider*: Nuisance Factors, Block Effects, and Batch Effects in Nonclinical Safety Assessment Studies

Detection of test article-related effects and the determination of the adversity of those changes are the primary goals of nonclinical safety assessment studies for drugs and chemicals in development. During these studies, variables that are not of primary interest to investigators may change and influence data interpretation. These variables, often referred to as "nuisance factors," may influence other groups of data and result in "block or batch effects" that complicate data interpretation. Definitions of the terms "nuisance factors," "block effects," and "batch effects," as they apply to nonclinical safety assessment studies, are reviewed. Multiple case examples of block and batch effects in safety assessment studies are provided, and the challenges these bring to pathology data interpretation are discussed. Methods to mitigate the occurrence of block and batch effects in safety assessment studies, including statistical blocking and utilization of study designs that minimize potential confounding variables, incorporation of adequate randomization, and use of an appropriate number of animals or repeated measurement of specific parameters for increased precision, are reviewed. [Box: see text].

The Tensile Performance of r-PET Bottle String: Effects of Different String Sizes and Cutting Axes

Recently, plastic waste has become one of the global issues which contributes to severe environmental pollution and threatens marine life as it takes a longer time to decompose. This study was aimed to protect the environment by recycling waste materials to reduce solid wastes. Recycled polyethylene terephthalate string was used in the study to identify tensile properties. The main aim of this research was to identify the tensile strength properties by investigating different string sizes for the r-PET bottle (4 mm, 5 mm, and 6 mm) at different cutting axes (horizontal and vertical). According to ASTM D882-02, the tensile strength test was conducted with a load cell of 50 N and crosshead speed of 50 mm/min at room temperature. To reduce the possibility of nuisance factors that will affect the result, five repetitive tests were carried out and Minitab R.18 software was used to analyse the obtained data. Results revealed that the 4 mm string in both cutting axes demonstrated the highest strength of tensile performance and modulus elasticity.

Using the difference in actual and expected calf liveweight relative to its dam liveweight as a statistic for interherd and intraherd benchmarking and genetic evaluations1.

The importance of improving the efficiency of beef production systems using both genetic and management strategies has long been discussed. Despite the contribution of the mature beef herd to the overall cost of production in the sector as a whole, most strategies for improving (feed) efficiency have focused on the growing animal. The objective of the present study was to quantify the phenotypic and genetic variability in several novel measures that relate the weight of a calf to that of its dam and vice versa. Two novel residual traits, representing the deviation in calf weight relative to its expectation from the population based on its dam's weight (DIFFcalf) or the deviation in the weight of the dam relative to its expectation from the population based on its calf's weight (DIFFdam), were calculated while simultaneously accounting for some nuisance factors in a multiple regression model. Four supplementary traits were also calculated, namely, 1) the deviation in calf weight from its expectation expressed relative to the weight of the dam (DIFFcalf_ratio), 2) the deviation in dam weight from its expectation relative to the weight of the dam (DIFFdam_ratio), 3) DIFFcalf-DIFFdam, and 4) the simple ratio of calf weight to its dam's weight (RATIOcalfdam). Genetic and residual variance components for each of the 6 traits were estimated using animal-dam linear mixed models. The phenotypic SD for DIFFcalf was 42 kg and, when expressed relative to the weight of the dam (i.e., DIFFcalf_ratio), was 0.07. The genetic SD for DIFFcalf and DIFFcalf_ratio was 16.66 kg and 0.02, respectively. The direct and maternal heritability estimated for DIFFcalf was 0.28 (SE = 0.04) and 0.11 (SE = 0.02), respectively, and for DIFFcalf_ratio was 0.24 (SE = 0.04) and 0.17 (SE = 0.03), respectively. The genetic SD for DIFFdam was 47.09 kg; the direct heritability was 0.50 (SE = 0.03), and the dam repeatability was 0.75 (SE = 0.01). The genetic SD for RATIOcalfdam was 0.03; the direct and maternal heritability was 0.24 (SE = 0.04) and 0.24 (SE = 0.03), respectively. The suggested traits outlined in the present study provide useful metrics for benchmarking dam-calf efficiency; in addition, the genetic variability detected in these traits suggest genetic progress for more efficient dam-calf pairs is indeed possible.

Nuisance effects in inter-scan functional connectivity estimates before and after nuisance regression

In resting-state functional MRI, the correlation between blood-oxygenation-level-dependent (BOLD) signals across brain regions is used to estimate the functional connectivity (FC) of the brain. FC estimates are prone to the influence of nuisance factors including scanner-related artifacts and physiological modulations of the BOLD signal. Nuisance regression is widely performed to reduce the effect of nuisance factors on FC estimates on a per-scan basis. However, a dedicated analysis of nuisance effects on the variability of FC metrics across a collection of scans has been lacking. This work investigates the effects of nuisance factors on the variability of FC estimates across a collection of scans both before and after nuisance regression. Inter-scan variations in FC estimates are shown to be significantly correlated with the geometric norms of various nuisance terms, including head motion measurements, signals derived from white-matter and cerebrospinal regions, and the whole-brain global signal (GS) both before and after nuisance regression. In addition, it is shown that GS regression (GSR) can introduce GS norm-related fluctuations that are negatively correlated with inter-scan FC estimates. The empirical results are shown to be largely consistent with the predictions of a theoretical framework previously developed for the characterization of dynamic FC measures. This work shows that caution must be exercised when interpreting inter-scan FC measures across scans both before and after nuisance regression.

Subspace techniques for task-independent EEG person identification.

There has been a growing interest in studying electroencephalography signals (EEG) as a possible biometric. The brain signals captured by EEG are rich and carry information related to the individual, tasks being performed, mental state, and other channel/measurement noise due to session variability and artifacts. To effectively extract person-specific signatures present in EEG, it is necessary to define a subspace that enhances the biometric information and suppresses other nuisance factors. i-vector and x-vector are state-of-art subspace techniques used in speaker recognition. In this paper, novel modifications are proposed for both frameworks to project person-specific signatures from multi-channel EEG into a subspace. The modified i-vector and x-vector systems outperform baseline i-vector and x-vector systems with an absolute improvement of 10.5% and 15.9%, respectively.

Curve description by histograms of tangent directions

The authors propose a novel approach for the description of objects based on contours in their images using real-valued feature vectors. The approach is particularly suitable when objects of interest have high contrast and texture-free images or when the texture variations are high so textural cues are nuisance factors for classification. The proposed descriptor is suitable for nearest neighbour classification still popular in embedded vision applications when the power considerations outweigh the performance requirements. They describe object outlines purely based on the histograms of contour tangent directions mimicking many of the design heuristics of texture-based descriptors such as scale-invariant feature transform (SIFT). However, unlike SIFT and its variants, the proposed approach is directly designed to work with contour data and it is robust to variations inside and outside the object outline as well as the sampling of the contour itself. They show that relying on tangent direction estimation as opposed to gradient computation yields a more robust description and higher nearest neighbour classification rates in a variety of classification problems.

Radial Acceleration Relation between Baryons and Dark or Phantom Matter in the Supercritical Acceleration Regime of Nearly Spherical Galaxies

Abstract The central regions of nearby elliptical galaxies are dominated by baryons (stars) and provide interesting laboratories for studying the radial acceleration relation (RAR). We carry out exploratory analyses and discuss the possibility of constraining the RAR in the supercritical acceleration range (10−9.5, 10−8) m s−2 by using a sample of nearly round pure-bulge (spheroidal, dispersion-dominated) galaxies. This sample includes 24 ATLAS3D galaxies and 4201 SDSS galaxies, and covers a wide range of masses, sizes, and luminosity density profiles. We consider a range of current possibilities for the stellar mass-to-light ratio (M ⋆/L), its gradient, and dark or phantom matter (DM/PM) halo profiles. We obtain the probability density functions (PDFs) of the parameters of the considered models via Bayesian inference based on spherical Jeans Monte Carlo modeling of the observed velocity dispersions. We then constrain the DM/PM-to-baryon acceleration ratio a X/a B from the PDFs. Unless we ignore observed radial gradients in M ⋆/L, or assume unreasonably strong gradients, marginalization over nuisance factors suggests with p = −1.00 ± 0.03 (stat) (sys) and q = −1.02 ± 0.09 (stat) (sys) around a supercritical acceleration a +1 ≡ 1.2 × 10−9 m s−2. In the context of the ΛCDM paradigm, this RAR suggests that the NFW DM halo profile is a reasonable description of galactic halos even after the processes of galaxy formation and evolution. In the context of the MOND paradigm, this RAR favors the “Simple” interpolating function but is inconsistent with the vast majority of other theoretical proposals and fitting functions motivated mainly by subcritical acceleration data.