Conventional Statistical Approaches Research Articles

Unsupervised Detection of Cell-Assembly Sequences by Similarity-Based Clustering.

Neurons which fire in a fixed temporal pattern (i.e., “cell assemblies”) are hypothesized to be a fundamental unit of neural information processing. Several methods are available for the detection of cell assemblies without a time structure. However, the systematic detection of cell assemblies with time structure has been challenging, especially in large datasets, due to the lack of efficient methods for handling the time structure. Here, we show a method to detect a variety of cell-assembly activity patterns, recurring in noisy neural population activities at multiple timescales. The key innovation is the use of a computer science method to comparing strings (“edit similarity”), to group spikes into assemblies. We validated the method using artificial data and experimental data, which were previously recorded from the hippocampus of male Long-Evans rats and the prefrontal cortex of male Brown Norway/Fisher hybrid rats. From the hippocampus, we could simultaneously extract place-cell sequences occurring on different timescales during navigation and awake replay. From the prefrontal cortex, we could discover multiple spike sequences of neurons encoding different segments of a goal-directed task. Unlike conventional event-driven statistical approaches, our method detects cell assemblies without creating event-locked averages. Thus, the method offers a novel analytical tool for deciphering the neural code during arbitrary behavioral and mental processes.

Drilling performance monitoring and optimization: a data-driven approach

Drilling performance monitoring and optimization are crucial in increasing the overall NPV of an oil and gas project. Even after rigorous planning, drilling phase of any project can be hindered by unanticipated problems, such as bit balling. The objective of this paper is to implement artificial intelligence technique to develop a smart model for more accurate and robust real-time drilling performance monitoring and optimization. For this purpose, the back propagation, feed forward neural network model was developed to predict rate of penetration (ROP) using different input parameters such as weight on bit, rotations per minute, mud flow (GPM) and differential pressures. The heavy hitter features identification and dimensionality reduction are performed to understand the impacts of each of the drilling parameters on ROP. This will be used to optimize the input parameters for model development and validation and performing the operation optimization when bit is underperforming. The model is first developed based on the drilling experiments performed in the laboratory and then extended to field applications. From both laboratory and field test data provided, we have proved that the data-driven model built using multilayer perceptron technique can be successfully used for drilling performance monitoring and optimization, especially identifying the bit malfunction or failure, i.e., bit balling. We have shown that the ROP has complex relationship with other drilling variables which cannot be captured using conventional statistical approaches or from different empirical models. The data-driven approach combined with statistical regression analysis provides better understanding of relationship between variables and prediction of ROP.

Dynamic multi-outcome prediction after injury: Applying adaptive machine learning for precision medicine in trauma.

ObjectiveMachine learning techniques have demonstrated superior discrimination compared to conventional statistical approaches in predicting trauma death. The objective of this study is to evaluate whether machine learning algorithms can be used to assess risk and dynamically identify patient-specific modifiable factors critical to patient trajectory for multiple key outcomes after severe injury.MethodsSuperLearner, an ensemble machine-learning algorithm, was applied to prospective observational cohort data from 1494 critically-injured patients. Over 1000 agnostic predictors were used to generate prediction models from multiple candidate learners for outcomes of interest at serial time points post-injury. Model accuracy was estimated using cross-validation and area under the curve was compared to select among predictors. Clinical variables responsible for driving outcomes were estimated at each time point.ResultsSuperLearner fits demonstrated excellent cross-validated prediction of death (overall AUC 0.94–0.97), multi-organ failure (overall AUC 0.84–0.90), and transfusion (overall AUC 0.87–0.9) across multiple post-injury time points, and good prediction of Acute Respiratory Distress Syndrome (overall AUC 0.84–0.89) and venous thromboembolism (overall AUC 0.73–0.83). Outcomes with inferior data quality included coagulopathic trajectory (AUC 0.48–0.88). Key clinical predictors evolved over the post-injury timecourse and included both anticipated and unexpected variables. Non-random missingness of data was identified as a predictor of multiple outcomes over time.ConclusionsMachine learning algorithms can be used to generate dynamic prediction after injury while avoiding the risk of over- and under-fitting inherent in ad hoc statistical approaches. SuperLearner prediction after injury demonstrates promise as an adaptable means of helping clinicians integrate voluminous, evolving data on severely-injured patients into real-time, dynamic decision-making support.

MACHINE LEARNING PREDICTION MODELS FOR IN-HOSPITAL MORTALITY AFTER TRANSCATHETER AORTIC VALVE IMPLANTATION AMONG RACIAL/ETHNIC MINORITY POPULATIONS IN THE UNITED STATES

Risk prediction tools for in-hospital complications in patients undergoing Transcatheter Aortic Valve Implantation (TAVI) are scarce. Current models have been designed using a conventional statistical approach with limited representation of minority groups. We sought to develop and compare an array

Bitcoin price forecasting with neuro-fuzzy techniques

Cryptocurrencies, with Bitcoin being the most notable example, have attracted considerable attention in recent years, and they have experienced large fluctuations in their price. While a few studies employ conventional statistical and econometric approaches to reveal the driving factors of Bitcoin's prices, research on the development of forecasting models to be used as decision support tools in investment strategies is scarce. This study proposes a computational intelligence technique that uses a hybrid Neuro-Fuzzy controller, namely PATSOS, to forecast the direction in the change of the daily price of Bitcoin. The proposed methodology outperforms two other computational intelligence models, the first being developed with a simpler neuro-fuzzy approach, and the second being developed with artificial neural networks. Furthermore, the investment returns achieved by a trading simulation, based on the signals of the proposed model, are 71.21% higher than the ones achieved through a naive buy-and-hold strategy. The performance of the PATSOS system is robust to the use of other cryptocurrencies.

Hydrogeochemical processes in the groundwater environment of Batlagundu block, Dindigul district, Tamil Nadu: conventional graphical and multivariate statistical approach

Water quality assessment is essential to ensure viable safe use of the resources for drinking, agricultural and industrial purposes. Deterioration of groundwater quality due to anthropogenic activities is increasing at an alarming rate in most parts of Dindigul district, but limited work has been carried out on groundwater quality and monitoring. This paper highlights the groundwater quality and compares its suitability for drinking and irrigation purpose of Batlagundu block in Dindigul district of Tamil Nadu. Eighteen groundwater samples were collected systematically in triplicates during the year 2013–2014. The physical and chemical parameters of the analytical results of groundwater were compared with the standard guideline values recommended by the Bureau of Indian Standards and World Health Organization for irrigational suitability, domestic usage. Further hydrogeochemical data have been plotted in standard graphs such as Piper trilinear diagram, US Salinity Laboratory (USSL), Giggenbach triangle diagram and Schoeller diagram. The mathematical models are used to estimate water quality parameters and to describe realistic water situations. It is proved that electrical conductance is an important water quality parameter. Hydrofacies diagram reported that alkaline earths (Ca2+ and Mg2+) significantly exceed the alkalis (Na+ and K+), while acids of (Cl− and HCO3−) exceed the acids of (SO42− and CO32−). Therefore, the chemical compositions of the study area were characterized by the Ca–Mg–HCO3 water type and Na–Cl type. USSL diagram of the study area revealed that most of the sampling sites possess high salinity and low-to-medium sodium hazard indicated that water samples are poorly suitable for usual agropurposes. Giggenbach triangular diagram indicated that the samples have not gained equilibrium with their host rocks, presumably due to fact circulation of fluid through the rock features. The present research work thus concludes that groundwater in the study area is chemically unsuitable for domestic and agricultural uses. It is recommended to carry out a continuous water quality monitoring program and development of effective management practices for utilization of water resources.

Development of strategy for simultaneous enhanced production of alkaline xylanase-pectinase enzymes by a bacterial isolate in short submerged fermentation cycle

The aim of this study is to enhance the production of industrially valuable xylanase and pectinase enzymes in short duration, using agrowaste extracted substrates. Conventional cum statistical multifactor analysis approaches were used in order to evaluate the effect of crude extracted substrates, supplemented for the production of xylanase-pectinase enzymes. Incorporation of crude extracted xylan (1.2 mg/ml of inoculum) and pectin (4.8 mg/ml of inoculum) substrates in inoculum resulted in maximal xylanase (320 ± 15) and pectinase titre (90 ± 8) after 48 h, using 2% wheat bran and 2% citrus peel in production medium with 48 h of fermentation time, with one variable factor at a time approach. The best condition obtained after performing statistical multifactor interaction analysis includes 5.50 mg/ml of pectin in inoculum,1.50 mg/ml of xylan in inoculum, wheat bran 3%, temperature 37.5 °C, time 48 h, 7 mg/ml of pectin in production medium, peptone 1.05%, inoculum size 2% and inoculum age of 20 h, with alkaline xylanase activity of 415.22 ± 18.50 IU/ml and alkaline pectinase activity of 109.10 ± 8.80 IU/ml. Activity of different pectinolytic enzymes per ml was also calculated, with 18.98 IU of exo-polymethylgalacturonase, 0.14 IU of endo-polymethylgalacturonase, 80 IU of exo-polygalacturonase, 0.28 IU of endo-polygalacturonase, 1.42 IU of polymethylgalacturonate lyase, 1.47 IU of polygalacturonate lyase, 0.15 IU of pectin esterase. This is the first report mentioning the utilization of crude extracted xylan and extracted pectin in inoculum to get the increment in the activity of both alkaline xylanase-pectinase enzymes simultaneously under short submerged fermentation cycle.

Machine-learning-based patient-specific prediction models for knee osteoarthritis.

Osteoarthritis (OA) is an extremely common musculoskeletal disease. However, current guidelines are not well suited for diagnosing patients in the early stages of disease and do not discriminate patients for whom the disease might progress rapidly. The most important hurdle in OA management is identifying and classifying patients who will benefit most from treatment. Further efforts are needed in patient subgrouping and developing prediction models. Conventional statistical modelling approaches exist; however, these models are limited in the amount of information they can adequately process. Comprehensive patient-specific prediction models need to be developed. Approaches such as data mining and machine learning should aid in the development of such models. Although a challenging task, technology is now available that should enable subgrouping of patients with OA and lead to improved clinical decision-making and precision medicine.

Artificial or intelligent? Machine learning and medical selection: possibilities and risks.

This article was migrated. The article was marked as recommended. Machine learning approaches form the basis of "artificial intelligence" and have been increasingly applied in health services settings. It has been shown that such approaches may produce more accurate predictions in some contexts, compared to conventional statistical approaches, and may also reduce the costs of decision-making through automation. Nevertheless, there are both general limitations to developing and implementing machine learning approaches that must be borne in mind. To date, relatively little research has been published on the potential for machine learning to support personnel selection. Moreover, there are particular challenges and issues that need to be considered if such methods are to be used to support decision-making in medical selection scenarios. This article describes some of these potential advantages and challenges and presents an illustrative example, based on real-world data, related to the selection of medical undergraduates.

Estimation of Clear-Water Local Scour at Pile Groups Using Genetic Expression Programming and Multivariate Adaptive Regression Splines

The physical process of scour around pile groups is complex. Due to economical and geotechnical considerations, multiple pile bridge piers have become more common in bridge designs. Various empirical models have been developed to estimate scour depth at pile groups. However, these models are mostly based on the conventional statistical regression approaches and are not able to adequately capture the highly nonlinear and complex relationship between scour depth and its influential factors. In this study, genetic expression programming (GEP) and multivariate adaptive regression splines (MARS) were utilized to estimate clear-water local scour depth at pile groups using the flow, sediment, and pile characteristics. Two combinations of data were used to train the GEP and MARS models. The first combination included dimensional variables (e.g., mean flow velocity and depth, mean grain diameter, pile diameter). The second combination contained nondimensional parameters. Results indicated that GEP and MARS can accurately estimate scour depth. Both models yielded better results when the dimensional data were used. In addition, the MARS model with a root mean square error (RMSE) of 0.0220 m and correlation coefficient (R2) of 0.902 outperformed the GEP model with an RMSE of 0.0285 m and R2 of 0.834. Performance of the GEP and MARS models was compared with that of the existing equations. The comparison showed that both models perform better than the regression-based empirical equations. Finally, a sensitivity analysis showed that pile diameter has the most significant impact on equilibrium scour depth.

Interactive reservoir computing for chunking information streams.

Chunking is the process by which frequently repeated segments of temporal inputs are concatenated into single units that are easy to process. Such a process is fundamental to time-series analysis in biological and artificial information processing systems. The brain efficiently acquires chunks from various information streams in an unsupervised manner; however, the underlying mechanisms of this process remain elusive. A widely-adopted statistical method for chunking consists of predicting frequently repeated contiguous elements in an input sequence based on unequal transition probabilities over sequence elements. However, recent experimental findings suggest that the brain is unlikely to adopt this method, as human subjects can chunk sequences with uniform transition probabilities. In this study, we propose a novel conceptual framework to overcome this limitation. In this process, neural networks learn to predict dynamical response patterns to sequence input rather than to directly learn transition patterns. Using a mutually supervising pair of reservoir computing modules, we demonstrate how this mechanism works in chunking sequences of letters or visual images with variable regularity and complexity. In addition, we demonstrate that background noise plays a crucial role in correctly learning chunks in this model. In particular, the model can successfully chunk sequences that conventional statistical approaches fail to chunk due to uniform transition probabilities. In addition, the neural responses of the model exhibit an interesting similarity to those of the basal ganglia observed after motor habit formation.

Visual gaze patterns reveal surgeons' ability to identify risk of bile duct injury during laparoscopic cholecystectomy

Introduction: Bile duct injury is the main serious complication of laparoscopic cholecystectomy. This study uses eye-tracking technology to identify distinct visual gaze patterns employed by surgeons able to promptly detect bile duct injury risk during laparoscopic cholecystectomy. Methods: Gaze tracking was performed using a remote eye-tracking system. Subjects were shown 3-minute videos of a laparoscopic cholecystectomy that led to a serious bile duct injury (Video A) and of an uneventful procedure (Video B), while logging if they perceived errors in surgical approach that could result in bile duct injury. Outcome parameters include fixation sequences on anatomical structures and eye tracking metrics. Participants were stratified based on performance and compared using conventional statistical and machine learning approaches. Results: 29 surgeons were recruited. There was no relationship between experience and speed of risk identification. The group with early-detection of the risk of injury displayed reduced dwell time [ms] on the common bile duct (6632.50 [IQR 3600.3] vs 9249.30 [IQR 3924.8]; p = 0.041) in the first half of video A. They also displayed increased cystic duct dwell time [ms] (15665.30 [IQR 7999.5] vs 12715.65 [8348.7]; p= 0.026) and increased Calot's triangle glances count (3.00 [IQR 2] vs 1.50 [IQR 1]; p = 0.018) viewing the dissection of Calot's triangle in video B. Hidden Markov Model based classification of fixation sequences demonstrated clear separability between early and late identifier groups. Conclusions: There are discernible differences in visual gaze patterns of surgeons during laparoscopic cholecystectomy associated with early recognition of impending bile duct injury.

Mixing regime-dependent causality between phytoplankton and bacteria in the subtropical North Atlantic Ocean ecosystem

MEPS Marine Ecology Progress Series Contact the journal Facebook Twitter RSS Mailing List Subscribe to our mailing list via Mailchimp HomeLatest VolumeAbout the JournalEditorsTheme Sections MEPS 600:41-53 (2018) - DOI: https://doi.org/10.3354/meps12643 Mixing regime-dependent causality between phytoplankton and bacteria in the subtropical North Atlantic Ocean ecosystem Hyewon Kim1,3,*, Dong Eun Lee2, Hugh W. Ducklow1 1Division of Biology and Paleo Environment, Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY 10964, USA 2Division of Ocean and Climate Physics, Lamont-Doherty Earth Observatory, Columbia University, Palisades, NY 10964, USA 3Present address: Department of Environmental Sciences, University of Virginia, Charlottesville, VA 22904, USA *Corresponding author: hk8m@virginia.edu ABSTRACT: Ever since marine heterotrophic bacteria were understood to mediate carbon fluxes to upper trophic levels via the microbial loop, quantifying the coupling between bacteria and phytoplankton has been one of the main priorities for marine microbiologists. The complex nature of phytoplankton-bacterial interactions may lead to nonlinear and regime-dependent coupling, for which conventional statistical approaches such as cross-correlation provide an incomplete picture of the dynamics. Here, we employed a nonlinear method for detecting causality, called convergent cross mapping (CCM), to examine a causal linkage between phytoplankton (primary production) and bacteria (bacterial production) at the Bermuda Atlantic Time-series Study (BATS) site. First, we verified the robustness of our CCM models with synthetic time series output from the Fasham-Ducklow-McKelvie ecosystem model. Initially, we hypothesized a strong bi-directional causal link between phytoplankton and bacteria due to the importance of the microbial loop at BATS. However, the results from our CCM models highlight that phytoplankton-bacterial causal coupling depends on seasonally distinct mixing regimes. While there was no evidence for bi-directional causality between phytoplankton and bacteria during the mixing period, moderate unidirectional causality of bacteria on phytoplankton was detected during the stratification period. Our study reveals causal associations between the 2 major microbial loop processes, for which better quantification is needed to improve our understanding of carbon cycling and export via microbial food webs. KEY WORDS: Convergent cross mapping · CCM · Bermuda Atlantic Time-series Study · BATS · Primary production · Bacterial production · Sargasso Sea Full text in pdf format Supplementary material PreviousNextCite this article as: Kim H, Lee DE, Ducklow HW (2018) Mixing regime-dependent causality between phytoplankton and bacteria in the subtropical North Atlantic Ocean ecosystem. Mar Ecol Prog Ser 600:41-53. https://doi.org/10.3354/meps12643 Export citation RSS - Facebook - Tweet - linkedIn Cited by Published in MEPS Vol. 600. Online publication date: July 30, 2018 Print ISSN: 0171-8630; Online ISSN: 1616-1599 Copyright © 2018 Inter-Research.

Complex Systems: Marketing's New Frontier

Complex systems approaches are emerging as new methods that complement conventional analytical and statistical approaches for analyzing marketing phenomena. These methods can provide researchers with tools to understand and predict marketing outcomes that emerge at the aggregate level by modeling feedback between heterogeneous agents and agent interaction with various marketing environmental variables. While the benefits of complex systems approaches often come with a high computational cost, steady advances in access to better computational resources has allowed more researchers to adopt complex systems approaches as part of their portfolio of methods. In this paper, we will provide a description of the key concepts, benefits, and tools of complex systems. The goal of this work is to encourage marketing researchers and practitioners who are not familiar with these approaches to consider the adoption of these methods. We end with a discussion of the future research opportunities that this powerful methodology enables.

Complex systems: marketing’s new frontier

Complex systems approaches are emerging as new methods that complement conventional analytical and statistical approaches for analyzing marketing phenomena. These methods can provide researchers with tools to understand and predict marketing outcomes that emerge at the aggregate level by modeling feedback between heterogeneous agents and agent interaction with various marketing environmental variables. While the benefits of complex systems approaches often come with a high computational cost, steady advances in access to better computational resources has allowed more researchers to adopt complex systems approaches as part of their portfolio of methods. In this paper, we will provide a description of the key concepts, benefits, and tools of complex systems. The goal of this work is to encourage marketing researchers and practitioners who are not familiar with these approaches to consider the adoption of these methods. We end with a discussion of the future research opportunities that this powerful methodology enables.

Heterogeneity between Pharmacoepidemiological Studies Using the Clinical Practice Research Datalink (CPRD)

The CPRD has been collecting medical data among UK patients since 1987. Many studies on the risk of cancer associated with prescription of antidiabetic drugs have used the CRPD database. The objective was to compare methods and results of studies on the same exposure and outcome combination based on the CPRD.A systematic Pubmed search of articles on diabetes and cancer based on the CPRD was performed. Articles were grouped into common exposure and outcome combinations. Methods and results were compared, especially in case of diverging results between studies, with a focus on study design and statistical methods. Forty-six studies were identified. Five of them studied colorectal cancer (CRC) occurrence in diabetic patients prescribed with metformin. All studies consistently found no significant increased or decreased CRC risk associated with metformin prescription. Five studies examined the risk of bladder cancer in relation to pioglitazone prescription. Two studies found a significantly increased risk of bladder cancer, one based on the analysis of the full dataset and one based on a nested case-control design using conventional matching procedures. The three other studies found no association, all of which used propensity score matched cohort design. Four studies investigated the risk of breast cancer among insulin glargine users. The three studies that included diabetic subjects who were prescribed glargine or another insulin therapy for the first time (i.e., the new user design) found no association. The study that included past and new users of glargine and of another insulin found a significantly increased risk. Despite the use of the same data source, studies on the same exposure and outcome may arrive at markedly different conclusions. Studies using propensity scores for matching subjects or including past users in cohorts may obtain results substantially different than studies using conventional statistical approaches. Disclosure A. Koechlin: None. P. Boyle: Other Relationship; Self; Sanofi. P. Autier: Other Relationship; Self; Sanofi.

Toward robust early-warning models: a horse race, ensembles and model uncertainty

This paper presents first steps towards robust models for crisis prediction. We conduct a horse race of conventional statistical methods and more recent machine learning methods as early-warning models. As individual models are in the literature most often built in isolation of other methods, the exercise is of high relevance for assessing the relative performance of a wide variety of methods. Further, we test various ensemble approaches to aggregating the information products of the built models, providing more robust basis for measuring country-level vulnerabilities. Finally, we provide approaches to estimating model uncertainty in early-warning exercises, particularly model performance uncertainty and model output uncertainty. The approaches put forward in this paper are shown with Europe as a playground. Generally, our results show that the conventional statistical approaches are outperformed by more advanced machine learning methods, such as k-nearest neighbours and neural networks, and particularly by model aggregation approaches through ensemble learning.

Machine learning for prediction of 30-day mortality after ST elevation myocardial infraction: An Acute Coronary Syndrome Israeli Survey data mining study

BackgroundRisk scores for prediction of mortality 30-days following a ST-segment elevation myocardial infarction (STEMI) have been developed using a conventional statistical approach. ObjectiveTo evaluate an array of machine learning (ML) algorithms for prediction of mortality at 30-days in STEMI patients and to compare these to the conventional validated risk scores. MethodsThis was a retrospective, supervised learning, data mining study. Out of a cohort of 13,422 patients from the Acute Coronary Syndrome Israeli Survey (ACSIS) registry, 2782 patients fulfilled inclusion criteria and 54 variables were considered.Prediction models for overall mortality 30days after STEMI were developed using 6 ML algorithms. Models were compared to each other and to the Global Registry of Acute Coronary Events (GRACE) and Thrombolysis In Myocardial Infarction (TIMI) scores. ResultsDepending on the algorithm, using all available variables, prediction models' performance measured in an area under the receiver operating characteristic curve (AUC) ranged from 0.64 to 0.91. The best models performed similarly to the Global Registry of Acute Coronary Events (GRACE) score (0.87 SD 0.06) and outperformed the Thrombolysis In Myocardial Infarction (TIMI) score (0.82 SD 0.06, p<0.05). Performance of most algorithms plateaued when introduced with 15 variables. Among the top predictors were creatinine, Killip class on admission, blood pressure, glucose level, and age. ConclusionsWe present a data mining approach for prediction of mortality post-ST-segment elevation myocardial infarction. The algorithms selected showed competence in prediction across an increasing number of variables. ML may be used for outcome prediction in complex cardiology settings.

North American paleoclimate reconstructions for the Last Glacial Maximum using an inverse modeling through iterative forward modeling approach applied to pollen data

AbstractThe inverse modeling through iterative forward modeling (IMIFM) approach was used to reconstruct Last Glacial Maximum (LGM) climates from North American fossil pollen data. The approach was validated using modern pollen data and observed climate data. While the large‐scale LGM temperature IMIFM reconstructions are similar to those calculated using conventional statistical approaches, the reconstructions of moisture variables differ between the two approaches. We used two vegetation models, BIOME4 and BIOME5‐beta, with the IMIFM approach to evaluate the effects on the LGM climate reconstruction of differences in water use efficiency, carbon use efficiency, and atmospheric CO2 concentrations. Although lower atmospheric CO2 concentrations influence pollen‐based LGM moisture reconstructions, they do not significantly affect temperature reconstructions over most of North America. This study implies that the LGM climate was very cold but not very much drier than present over North America, which is inconsistent with previous studies.

Corporate environmental sustainability reporting in the context of national cultures: A quantile regression approach

This work, based on previous evidence within international business management research, aims to test the influence of communities’ culture features on corporate environmental sustainability reporting (CESR) practices. To overcome some limitations of conventional statistical approaches applied by previous research, a quantile regression (QR) model is implemented which allows setting a framework to test the working hypotheses in different scenarios that cover divergent firms’ commitment levels to stakeholder engagement and CESR practices development. Our central results addresses that different national culture dimensions present a non-monotonic influence on CESR practices. This result, which is analyzed through the Stakeholder Theory proposals, can be explained because corporate sustainability behaviors are highly sensitive to stakeholders’ pressures and demands which are ultimately conditioned by the cultural environment. Some interesting recommendations for companies’ strategic management and governmental policy-making processes are reported.