Modern Statistical Approaches Research Articles

Statistical Analysis Applied to the Production of Mirto Liqueur

Preparation of myrtle liqueur through ethanol-based extraction is a widely employed methodology. Nevertheless, optimization of existing processes is possible, especially through a modern statistical multivariate approach. In this context, a Design of Experiments (DoE) approach was used to quantitatively assess for the first time the effect of the time, ethanol concentration, temperature, and the ratio between the Myrtus communis berries’ weight and the extractant volume (v/w) on the amounts of anthocyanins, volatile compounds and dry residues in the liqueur. The kinetic profile relative to the volatile fraction variation during the process was described by gas chromatography (GC), while spectrophotometric analysis allowed quantification of the total anthocyanins and total polyphenols. Multiple response analysis showed that the maximum efficiencies in terms of the considered parameters (desirability function) were reached by setting the temperature to 25 °C and the ethanol percentage to 96% after 20 days of processing. Some hints as to the chemical instability and not negligible sensitivity of anthocyanins in relation to the experimental conditions for longer extraction times were also observed. The statistical model represents a novel tool for industrial production of myrtle liqueur.

MSR58 Comparing Predictive Models: Traditional Versus Modern Statistical Approaches for Economic Outcomes in Major Depressive Disorder

MSR58 Comparing Predictive Models: Traditional Versus Modern Statistical Approaches for Economic Outcomes in Major Depressive Disorder

New statistical framework for interlaboratory evaluation of anti-doping testing results by WADA

AbstractThe World Anti-doping Agency (WADA) International Standard for Laboratories (ISL), developed as part of the World Anti-Doping Program, requires satisfactory laboratory performance in the WADA External Quality Assessment Scheme (EQAS) in order to obtain and maintain WADA accreditation. Under this mandate, WADA regularly distributes urine and blood test samples to anti-doping laboratories to continuously monitor their proficiency. Over the years, WADA has employed classical, generic statistical methods, in accordance to ISO 13528, to evaluate quantitative EQAS results. Here, we set out the rationale for a modern statistical approach that recognizes and addresses the particular features of the measurement results typically obtained in such tests and present an approach involving Bayesian measurement models and statistical data analysis that is tailored specifically to anti-doping testing.

Further Clarification of Pain Management Complexity in Radiotherapy: Insights from Modern Statistical Approaches.

The primary objective of this study was to assess the adequacy of analgesic care in radiotherapy (RT) patients, with a secondary objective to identify predictive variables associated with pain management adequacy using a modern statistical approach, integrating the Least Absolute Shrinkage and Selection Operator (LASSO) algorithm and the Classification and Regression Tree (CART) analysis. This observational, multicenter cohort study involved 1387 patients reporting pain or taking analgesic drugs from 13 RT departments in Italy. The Pain Management Index (PMI) served as the measure for pain control adequacy, with a PMI score < 0 indicating suboptimal management. Patient demographics, clinical status, and treatment-related factors were examined to discern the predictors of pain management adequacy. Among the analyzed cohort, 46.1% reported inadequately managed pain. Non-cancer pain origin, breast cancer diagnosis, higher ECOG Performance Status scores, younger patient age, early assessment phase, and curative treatment intent emerged as significant determinants of negative PMI from the LASSO analysis. Notably, pain management was observed to improve as RT progressed, with a greater discrepancy between cancer (33.2% with PMI < 0) and non-cancer pain (73.1% with PMI < 0). Breast cancer patients under 70 years of age with non-cancer pain had the highest rate of negative PMI at 86.5%, highlighting a potential deficiency in managing benign pain in younger patients. The study underscores the dynamic nature of pain management during RT, suggesting improvements over the treatment course yet revealing specific challenges in non-cancer pain management, particularly among younger breast cancer patients. The use of advanced statistical techniques for analysis stresses the importance of a multifaceted approach to pain management, one that incorporates both cancer and non-cancer pain considerations to ensure a holistic and improved quality of oncological care.

Linear mixed-effect modeling of organ of Corti vibratory tuning curves

Optical coherence tomography has become the most popular approach to experimental measures of sound-induced vibrations within the mammalian cochlea. Because it is relatively easy to use and works in the unopened cochlea, the measurement of vibratory tuning curves has become highly reliable, and averaging data from multiple animals in different experimental cohorts is now possible. Here I tested a modern statistical approach to compare cohorts for differences in the magnitude and phase of vibration. A linear mixed-effect approach with first, second, third, and fourth-order models to fit the data was tested. The third-order model best fit both the magnitude and phase data without having terms that did not contribute substantively to improving the R2 or the p-value for the independent variables. It identified a difference between cohorts of mice that were different and no difference between cohorts that should not be different. Thus, this approach provides a way to simply compare a full set of tuning curves between cohorts. While further analyses by the investigator will always be needed to study specific details related to the study hypothesis, this statistical technique provides a simple way for the cochlear physiologist to perform an initial assessment of whether the cohorts are same or different.

Overview of modern approaches for identifying and evaluating heterogeneous treatment effects from clinical data.

There has been much interest in the evaluation of heterogeneous treatment effects (HTE) and multiple statistical methods have emerged under the heading of personalized/precision medicine combining ideas from hypothesis testing, causal inference, and machine learning over the past 10-15 years. We discuss new ideas and approaches for evaluating HTE in randomized clinical trials and observational studies using the features introduced earlier by Lipkovich, Dmitrienko, and D'Agostino that distinguish principled methods from simplistic approaches to data-driven subgroup identification and estimating individual treatment effects and use a case study to illustrate these approaches. We identified and provided a high-level overview of several classes of modern statistical approaches for personalized/precision medicine, elucidated the underlying principles and challenges, and compared findings for a case study across different methods. Different approaches to evaluating HTEs may produce (and actually produced) highly disparate results when applied to a specific data set. Evaluating HTE with machine learning methods presents special challenges since most of machine learning algorithms are optimized for prediction rather than for estimating causal effects. An additional challenge is in that the output of machine learning methods is typically a "black box" that needs to be transformed into interpretable personalized solutions in order to gain acceptance and usability.

God (≡ Elohim), The First Small World Network

In this article, the approach of network mapping of words in literary texts is extended to “textual factors”: the network nodes are defined as “concepts”; the links are “community connexions.” Thereafter, the text network properties are investigated along modern statistical physics approaches of networks, thereby relating network topology and algebraic properties to literary text contents. As a practical illustration, the first chapter of Genesis in the Bible is mapped into a 10-node network, as in the Kabbalah approach, mentioning God (≡ Elohim). The characteristics of the network are studied starting from its adjacency matrix and the corresponding Laplacian matrix. Triplets of nodes are particularly examined in order to emphasize the “textual (community) connexions” of each agent “emanation,” through the so-called clustering coefficients and the overlap index, hence measuring the “semantic flow” between the different nodes. It is concluded that this graph is a small world network and weakly dis-assortative, because its average local clustering coefficient is significantly higher than a random graph constructed on the same vertex set.

A Better Mechanistic Understanding of Big Data through an Order Search Using Causal Bayesian Networks

Every year, biomedical data is increasing at an alarming rate and is being collected from many different sources, such as hospitals (clinical Big Data), laboratories (genomic and proteomic Big Data), and the internet (online Big Data). This article presents and evaluates a practical causal discovery algorithm that uses modern statistical, machine learning, and informatics approaches that have been used in the learning of causal relationships from biomedical Big Data, which in turn integrates clinical, omics (genomic and proteomic), and environmental aspects. The learning of causal relationships from data using graphical models does not address the hidden (unknown or not measured) mechanisms that are inherent to most measurements and analyses. Also, many algorithms lack a practical usage since they do not incorporate current mechanistic knowledge. This paper proposes a practical causal discovery algorithm using causal Bayesian networks to gain a better understanding of the underlying mechanistic process that generated the data. The algorithm utilizes model averaging techniques such as searching through a relative order (e.g., if gene A is regulating gene B, then we can say that gene A is of a higher order than gene B) and incorporates relevant prior mechanistic knowledge to guide the Markov chain Monte Carlo search through the order. The algorithm was evaluated by testing its performance on datasets generated from the ALARM causal Bayesian network. Out of the 37 variables in the ALARM causal Bayesian network, two sets of nine were chosen and the observations for those variables were provided to the algorithm. The performance of the algorithm was evaluated by comparing its prediction with the generating causal mechanism. The 28 variables that were not in use are referred to as hidden variables and they allowed for the evaluation of the algorithm’s ability to predict hidden confounded causal relationships. The algorithm’s predicted performance was also compared with other causal discovery algorithms. The results show that incorporating order information provides a better mechanistic understanding even when hidden confounded causes are present. The prior mechanistic knowledge incorporated in the Markov chain Monte Carlo search led to the better discovery of causal relationships when hidden variables were involved in generating the simulated data.

Open-source analytical pipeline for robust data analysis, visualizations and sharing in crop breeding

BackgroundDeveloping a systematic phenotypic data analysis pipeline, creating enhanced visualizations, and interpreting the results is crucial to extract meaningful insights from data in making better breeding decisions. Here, we provide an overview of how the Rainfed Rice Breeding (RRB) program at IRRI has leveraged R computational power with open-source resource tools like R Markdown, plotly, LaTeX, and HTML to develop an open-source and end-to-end data analysis workflow and pipeline, and re-designed it to a reproducible document for better interpretations, visualizations and easy sharing with collaborators.ResultsWe reported the state-of-the-art implementation of the phenotypic data analysis pipeline and workflow embedded into a well-descriptive document. The developed analytical pipeline is open-source, demonstrating how to analyze the phenotypic data in crop breeding programs with step-by-step instructions. The analysis pipeline shows how to pre-process and check the quality of phenotypic data, perform robust data analysis using modern statistical tools and approaches, and convert it into a reproducible document. Explanatory text with R codes, outputs either in text, tables, or graphics, and interpretation of results are integrated into the unified document. The analysis is highly reproducible and can be regenerated at any time. The analytical pipeline source codes and demo data are available at https://github.com/whussain2/Analysis-pipeline.ConclusionThe analysis workflow and document presented are not limited to IRRI’s RRB program but are applicable to any organization or institute with full-fledged breeding programs. We believe this is a great initiative to modernize the data analysis of IRRI’s RRB program. Further, this pipeline can be easily implemented by plant breeders or researchers, helping and guiding them in analyzing the breeding trials data in the best possible way.

Alternative Data in Investment Management: Usage, Challenges, and Valuation

<i>Alternative data</i> in finance is an umbrella term for diverse nontraditional datasets used by quantitative and fundamental institutional investors to enhance portfolio returns. Although the use of alternative data is a recent phenomenon, it was not until the last five years that it gained widespread acceptance and the sector started to evolve into a complex ecosystem of data originators, intermediaries, and investors. The alternative data industry faces several obstacles, including difficulty estimating a dataset’s value to investors and technical challenges in leveraging these datasets efficiently at a large scale. In this article, the authors provide an up-to-date description of the alternative data space as it relates to the institutional investment industry. The authors elaborate on what alternative data are and how they are used in investment management. The authors identify and discuss some of the key challenges that arise when working with alternative data. In particular, they address issues such as entity mapping, ticker-tagging, panel stabilization, and debiasing with modern statistical and machine learning approaches. The authors advance several methodologies for the valuation of alternative datasets, including an event study methodology they refer to as the <i>golden triangle</i>, the application of report cards, and the relationship between a dataset’s structure of information content and its potential to enhance investment returns. To illustrate the effectiveness of these methods, they apply them to a case study analysis of real-world healthcare data, delivering an improvement in revenue prediction accuracy from an 88% mean absolute error to a 2.6% mean absolute error. <b>Key Findings</b> ▪ The authors describe the alternative data industry in an investment context. They discuss the commercial incentives for different categories of organizations within the industry and methodologies to valuate alternative datasets. ▪ They identify key technical challenges in working with alternative data, including entity mapping, ticker-tagging, panel stabilization, and debiasing. ▪ Through a case study, they demonstrate how to use alternative data combined with imputation and optimization techniques to predict revenues of publicly traded companies within the healthcare sector.

Kin Against Kin: Internal Co-selection and the Coherence of Kinship Typologies

Across the world people in different societies structure their family relationships in many different ways. These relationships become encoded in their languages as kinship terminology, a word set that maps variably onto a vast genealogical grid of kinship categories, each of which could in principle vary independently. But the observed diversity of kinship terminology is considerably smaller than the enormous theoretical design space. For the past century anthropologists have captured this variation in typological schemes with only a small number of model system types. Whether those types exhibit the internal co-selection of parts implicit in their use is an outstanding question, as is the sufficiency of typologies in capturing variation as a whole. We interrogate the coherence of classic kinship typologies using modern statistical approaches and systematic data from a new database, Kinbank. We first survey the canonical types and their assumed patterns of internal and external co-selection, then present two data-driven approaches to assess internal coherence. Our first analysis reveals that across parents’ and ego’s (one’s own) generation, typology has limited predictive value: knowing the system in one generation does not reliably predict the other. Though we detect limited co-selection between generations, “disharmonic” systems are equally common. Second, we represent structural diversity with a novel multidimensional approach we term kinship space. This approach reveals, for ego’s generation, some broad patterning consistent with the canonical typology, but diversity (and mixed systems) is considerably higher than classical typologies suggest. Our results strongly challenge the descriptive adequacy of the set of canonical kinship types.

Tracing molecular properties throughout evolution: A chemoinformatic approach

Evolution of metabolism is a longstanding yet unresolved question, and several hypotheses were proposed to address this complex process from a Darwinian point of view. Modern statistical bioinformatic approaches targeted to the comparative analysis of genomes are being used to detect signatures of natural selection at the gene and population level, as an attempt to understand the origin of primordial metabolism and its expansion. These studies, however, are still mainly centered on genes and the proteins they encode, somehow neglecting the small organic chemicals that support life processes. In this work, we selected steroids as an ancient family of metabolites widely distributed in all eukaryotes and applied unsupervised machine learning techniques to reveal the traits that natural selection has imprinted on molecular properties throughout the evolutionary process. Our results clearly show that sterols, the primal steroids that first appeared, have more conserved properties and that, from then on, more complex compounds with increasingly diverse properties have emerged, suggesting that chemical diversification parallels the expansion of biological complexity. In a wider context, these findings highlight the worth of chemoinformatic approaches to a better understanding the evolution of metabolism.

Unbiased and efficient log-likelihood estimation with inverse binomial sampling.

The fate of scientific hypotheses often relies on the ability of a computational model to explain the data, quantified in modern statistical approaches by the likelihood function. The log-likelihood is the key element for parameter estimation and model evaluation. However, the log-likelihood of complex models in fields such as computational biology and neuroscience is often intractable to compute analytically or numerically. In those cases, researchers can often only estimate the log-likelihood by comparing observed data with synthetic observations generated by model simulations. Standard techniques to approximate the likelihood via simulation either use summary statistics of the data or are at risk of producing substantial biases in the estimate. Here, we explore another method, inverse binomial sampling (IBS), which can estimate the log-likelihood of an entire data set efficiently and without bias. For each observation, IBS draws samples from the simulator model until one matches the observation. The log-likelihood estimate is then a function of the number of samples drawn. The variance of this estimator is uniformly bounded, achieves the minimum variance for an unbiased estimator, and we can compute calibrated estimates of the variance. We provide theoretical arguments in favor of IBS and an empirical assessment of the method for maximum-likelihood estimation with simulation-based models. As case studies, we take three model-fitting problems of increasing complexity from computational and cognitive neuroscience. In all problems, IBS generally produces lower error in the estimated parameters and maximum log-likelihood values than alternative sampling methods with the same average number of samples. Our results demonstrate the potential of IBS as a practical, robust, and easy to implement method for log-likelihood evaluation when exact techniques are not available.

Alternative Data in Investment Management: Usage, Challenges and Valuation

Alternative data in finance is an umbrella term for diverse non-traditional datasets used by quantitative and fundamental institutional investors to enhance portfolio returns. While the use of alternative data is a recent phenomenon, it was not until the last five years that it gained widespread acceptance and the sector started to evolve into a complex ecosystem of data originators, intermediaries and investors. The alternative data industry faces several obstacles, including difficulty estimating a dataset’s value to investors and technical challenges in leveraging these datasets efficiently at large scale. In this article, the authors provide an up-to-date description of the alternative data space as it relates to the institutional investment industry. The authors elaborate on what alternative data is and how it is used in investment management. The authors identify and discuss some of the key challenges that arise when working with alternative data. In particular, they address issues such as entity mapping, ticker-tagging, panel stabilization and debiasing with modern statistical and machine learning approaches. The authors advance several methodologies for the valuation of alternative datasets, including an event study methodology they refer to as the Golden Triangle, the application of report cards, and the relationship between a datasets’ structure of information content its potential to enhance investment returns. To illustrate the effectiveness of the methods, they apply them to a case study analysis of real-world healthcare data, delivering an improvement of revenue prediction accuracy from an 88% mean absolute error to a 2.6% mean absolute error.

Modelling Autonomous Agents’ Decisions in Learning to Cross a Cellular Automaton-Based Highway via Artificial Neural Networks

A lot of effort has been devoted to mathematical modelling and simulation of complex systems for a better understanding of their dynamics and control. Modelling and analysis of computer simulations outcomes are also important aspects of studying the behaviour of complex systems. It often involves the use of both traditional and modern statistical approaches, including multiple linear regression, generalized linear model and non-linear regression models such as artificial neural networks. In this work, we first conduct a simulation study of the agents’ decisions learning to cross a cellular automaton based highway and then, we model the simulation data using artificial neural networks. Our research shows that artificial neural networks are capable of capturing the functional relationships between input and output variables of our simulation experiments, and they outperform the classical modelling approaches. The variable importance measure techniques can consistently identify the most dominant factors that affect the response variables, which help us to better understand how the decision-making by the autonomous agents is affected by the input factors. The significance of this work is in extending the investigations of complex systems from mathematical modelling and computer simulations to the analysis and modelling of the data obtained from the simulations using advanced statistical models.

INVESTIGATING ECOLOGICAL AND PHYLOGENETIC CONSTRAINTS IN HIPPOPOTAMIDAE SKULL SHAPE

Hippopotamidae are a group of large-sized mammals of interest for testing evolutionary traits in time and space. Variation in skull shape within Hippopotamidae is here investigated by means of shape analysis (Geometric Morphometrics) and modern statistical approaches. Two-dimensional shape analysis is applied to dorsal and lateral views of extant and extinct Hippopotamidae species sufficiently preserved to allow their morphology to be captured by landmark and semi-landmark digitization. The results show that Hippopotamus gorgops and H. antiquus display similar shapes, while Hexaprotodon palaeindicus falls within the morphospace occupied by H. amphibius, suggesting similar morphology. The cranial shape of the Sicilian hippopotamus (H. pentlandi) still resembles that of H. amphibius in lateral view, suggesting that adaptation to the insular domain was yet not fully attained. Madagascan hippopotamuses (H. madagascariensis and H. lemerlei) are close to the pygmy hippo, Choeropsis liberiensis, in PC1 values; nevertheless, the cranial shape of the Madagascan hippos seems not to be closely related to the cranial shape of C. liberiensis. Despite the morphological convergences within the group, while cranial shape in Hippopotamidae is phylogenetically structured, this does not hold for size. Although further investigations are needed to test the influence of ecological and palaeoecological parameters on the general shape to provide additional information for understanding Hippopotamidae evolution and adaptation, the present study provides an insight into the evolutionary framework of Hippopotamidae.

A Comparison of Flare Forecasting Methods. III. Systematic Behaviors of Operational Solar Flare Forecasting Systems

Abstract A workshop was recently held at Nagoya University (2017 October 31–November 2), sponsored by the Center for International Collaborative Research, at the Institute for Space-Earth Environmental Research, Nagoya University, Japan, to quantitatively compare the performance of today’s operational solar flare forecasting facilities. Building upon Paper I of this series, in Paper II we described the participating methods for this latest comparison effort, the evaluation methodology, and presented quantitative comparisons. In this paper, we focus on the behavior and performance of the methods when evaluated in the context of broad implementation differences. Acknowledging the short testing interval available and the small number of methods available, we do find that forecast performance: (1) appears to improve by including persistence or prior flare activity, region evolution, and a human “forecaster in the loop”; (2) is hurt by restricting data to disk-center observations; (3) may benefit from long-term statistics but mostly when then combined with modern data sources and statistical approaches. These trends are arguably weak and must be viewed with numerous caveats, as discussed both here and in Paper II. Following this present work, in Paper IV (Park et al. 2019) we will present a novel analysis method to evaluate temporal patterns of forecasting errors of both types (i.e., misses and false alarms). Hence, most importantly, with this series of papers, we demonstrate the techniques for facilitating comparisons in the interest of establishing performance-positive methodologies.

Rasch model of the Child Perceptions Questionnaire for oral health–related quality of life: A step forward toward accurate outcome measures

BackgroundOral health is related to major chronic diseases and contributes to general health. Oral health–related quality of life (OHRQoL) is, therefore, an essential part of well-being. Measurement of OHRQoL over the lifetime requires accurate and comparable outcome measures that are suitable for different age groups. The most frequently used instrument for children aged 11 through 14 years is the Child Perceptions Questionnaire (CPQ-11-14). The authors aimed to examine the psychometric properties of the CPQ-11-14 using modern statistical approaches. MethodsThe authors carried out a psychometric analysis of 2 population-representative data sets from Germany (N = 2,658) and a randomly selected smaller sample (n = 300) on the basis of the Rasch model. ResultsThe authors detected a right-skewed distribution of the CPQ-11-14 total scores. The instrument items did not match the level of OHRQoL of the study participants. Moreover, a large number of questionnaire items (range, 51.4%-94.3%) had so-called disordered thresholds, meaning that the answer categories did not work as intended. Unidimensionality was achieved via splitting the CPQ-11-14 into 2 subscales, namely socioemotional and oral symptoms and functioning. ConclusionsDespite its classic validation, the CPQ-11-14 needs adaptions for meaningful clinical use. More items on the better end of the scale are needed to better differentiate between children with higher and lower OHRQoL. Practical ImplicationsThe Rasch model addresses psychometric properties of questionnaires on an item-based level, which were not detected via classic approaches. The results reported by the authors are needed to generate a well-calibrated outcome measure for accurate clinical and public health assessment of OHRQoL.

Bayesian estimation of total fertility from a population's age–sex structure

We investigate a modern statistical approach to a classic deterministic demographic estimation technique. When vital event registration is missing or inadequate, it is possible to approximate a population's total fertility rate (TFR) from information about its distribution by age and sex. For example, if under-five child mortality is low then TFR is often close to seven times the child/woman ratio (CWR), the number of 0–4 year olds per 15–49-year-old woman. We analyse the formal relationship between CWR and TFR to identify sources of uncertainty in indirect estimates. We construct a Bayesian model for the statistical distribution of TFR conditional on the population's age–sex structure, in which unknown demographic quantities in the standard approximation are parameters with prior distributions. We apply the model in two case studies: to a small indigenous population in the Amazon region of Brazil that has extremely high fertility rates, and to the set of 159 counties in the US state of Georgia. A statistical approach yields important insights into the sources of error in indirect estimation, and their relative magnitudes.

Digital rock models of Precambrian and Paleozoic tight formations from Poland

Properties of selected Precambrian and Paleozoic sedimentary clastic rocks were analysed with respect to their reservoir potential. Multidimensional analysis of laboratory results and borehole logging data was used to construct digital models of pre-Mesozoic, deeply buried formations, present as tight, low-porosity and low-permeability rocks. This modern statistical and deterministic approach as applied to laboratory and borehole logging results worked to integrate data at different scales. The results obtained are useful not only in further scientific research but also found a use in industrial application. As a first step, statistical methods, including clustering and separation of homogeneous groups, enabled digital rock model creation on the basis of the results of such laboratory measurements as pycnometry, mercury porosimetry, nuclear magnetic resonance spectroscopy or computed X-ray tomography. Next, the models constructed were applied in borehole logging interpretation to find intervals with similar petrophysical properties within the group and different properties between the groups. This approach allowed implementation of upscaling procedures of laboratory experiments at micro- and nano-scale to borehole logging scale. High correlations were established between the log petrophysical parameters within the digital models. This approach can be used to divide the succession cored into intervals with different petrophysical parameters.