Model Building Process Research Articles

A new approach to formulate fuzzy regression models

A fuzzy regression model is developed to construct the relationship between the response and explanatory variables in fuzzy environments. To enhance explanatory power and take into account the uncertainty of the formulated model and parameters, a new operator, called the fuzzy product core (FPC), is proposed for the formulation processes to establish fuzzy regression models with fuzzy parameters using fuzzy observations that include fuzzy response and explanatory variables. In addition, the sign of parameters can be determined in the model-building processes. Compared to existing approaches, the proposed approach reduces the amount of unnecessary or unimportant information arising from fuzzy observations and determines the sign of parameters in the models to increase model performance. This improves the weakness of the relevant approaches in which the parameters in the models are fuzzy and must be predetermined in the formulation processes. The proposed approach outperforms existing models in terms of distance, mean similarity, and credibility measures, even when crisp explanatory variables are used.

PNS297 CLINICAL AND KEY OPINION LEADERS VALIDATION IN HEALTH ECONOMICS MODEL AND COMPARATIVE EFFECTIVENESS: A GUIDE FOR NON-EXPERT AUDIENCE

Clinical validation — or face validation — is considered necessary in health economic (HE) modeling and comparative effectiveness analysis. Evaluation of model assumptions and interpretations by clinical experts provides an estimation of the clinical relevance of HE analysis. No specific framework including timing for this critical, and potentially costly and time-consuming consultation is available. Here we elaborate a framework, including timing for HE model clinical validation. A targeted literature review was conducted to highlight main themes and issues involved in establishing a conceptual framework for time- and cost-effective clinical validation. Results from the literature review emphasized the importance of clinical validation by clinical experts during three main phases of building a HE model: model structure, inputs and assumptions, and results. It was highlighted that face validity should be started early and iteratively within the analysis timeframe; however, no specific schedule or timing has yet been published. We suggest clinical validation in the process of HE model design should take place in three different phases. First, at the initial model design stage to provide clinical expertise about the underlying disease, patients, and treatment pathway. A second clinical consultation following initial efforts at providing model inputs and assumption should also be considered. A final clinical validation should be scheduled when results from the model are available. We recommend three or five clinical experts to be consulted by a modified Delphi panel with three rounds of emails — corresponding to the three phases described — including specific questions, but also utilizing targeted interviews when time and resources permit. Payers and economists commonly accept that clinical validation should be integral to the process of HE model-building. Timing of clinical validation is key to deriving full value to assist in conceptual framework design and filling data gaps.

Integration of Chlorpyrifos Acetylcholinesterase Inhibition, Water Temperature, and Dissolved Oxygen Concentration into a Regional Scale Multiple Stressor Risk Assessment Estimating Risk to Chinook Salmon

We estimated the risk to populations of Chinook salmon (Oncorhynchus tshawytscha) due to chlorpyrifos (CH), water temperature (WT), and dissolved oxygen concentration (DO) in 4 watersheds in Washington State, USA. The watersheds included the Nooksack and Skagit Rivers in the Northern Puget Sound, the Cedar River in the Seattle-Tacoma corridor, and the Yakima River, a tributary of the Columbia River. The Bayesian network relative risk model (BN-RRM) was used to conduct this ecological risk assessment and was modified to contain an acetylcholinesterase (AChE) inhibition pathway parameterized using data from CH toxicity data sets. The completed BN-RRM estimated risk at a population scale to Chinook salmon employing classical matrix modeling runs up to 50-y timeframes. There were 3 primary conclusions drawn from the model-building process and the risk calculations. First, the incorporation of an AChE inhibition pathway and the output from a population model can be combined with environmental factors in a quantitative fashion. Second, the probability of not meeting the management goal of no loss to the population ranges from 65% to 85%. Environmental conditions contributed to a larger proportion of the risk compared to CH. Third, the sensitivity analysis describing the influence of the variables on the predicted risk varied depending on seasonal conditions. In the summer, WT and DO were more influential than CH. In the winter, when the seasonal conditions are more benign, CH was the driver. Fourth, in order to reach the management goal, we calculated the conditions that would increase juvenile survival, adult survival, and a reduction in toxicological effects. The same process in this example should be applicable to the inclusion of multiple pesticides and to more descriptive population models such as those describing metapopulations. Integr Environ Assess Manag 2019;00:1-15. © 2019 SETAC.

Single Order of Multiple Regression Model of Water Quality Index (WQI) in Manjung River and its Tributaries

This research highlights a multi-variety technique to examine the relationship between dependent and independent variable in predicting the water quality index in Manjung Rivers and its affluents. The model building process been used to analyse and generate the data. There are 63 possible models for single order multiple regressions. The number of possible model started to reduce as we started to eliminate insignificant variable. This model then needs to run under eight selection criteria to identify the best model. The best model will be certified by using Mean absolute percentage error (MAPE) in order to measure the validity of the model.

Modelling of the propulsion plant of a large container ship by the partition of the cycle of the main diesel engine

In this article, a mathematical replica of the propulsive installation of a large container ship is presented. The ship propulsion is accomplished by a large two-stroke marine diesel engine driving a marine propeller. The main new idea introduced by this research consists in using the Subsystem-Enabling feature available in the MATLAB Simulink® environment to control the execution of the working sequences of the main propulsion two-stroke diesel engine. The benefits brought by this model implementation approach are its simplicity and its applicability to all the blocks of the diesel engine model, these blocks can be created to represent one engine working sequence and then duplicated to represent the remaining engine sequences, and finally, the blocks related to each sequence can be grouped in subsystems and controlled by a single subsystem monitoring the engine events according to the value of the crankshaft angle; consequently, the overall Simulink model building and execution processes are much faster. The results generated by the simulation using the Simulink model of this diesel ship propulsion plant show that, for the production of acceptable results, only a small portion of the input data of the model needs to be exactly provided by the manufacturers of the components of the propulsion plant, whereas the rest of the data can be given reasonable initial values and adjusted during the model’s fine-tuning process. Finally, it is worth noting that the considerable amount of simulator-input-data available in this article can be used in any other simulation work to develop an identical simulator to the simulator of this article or to build a different one from scratch.

Гібридний підхід до прогнозування часового ряду електроспоживання для орга- нізаційного управління на оптовому ринку

The problem of increasing the efficiency of solving the complex of tasks of forecasting and planning electric consumption by regional companies of electricity suppliers subjects of the organizational management system in the wholesale electricity market is considered. The analysis of the use of various modeling methods in solving the problem of choosing and building a model for forecasting electric consumption is carried out. The task of constructing a hybrid prognostic model devoid of the shortcomings of individual modeling methods is formulated. Preference is given to the approach associated with the integrated use of mathematical tools based on apparatus of artificial neural networks, a genetic algorithm and a Kalman filter for constructing generalized nonlinear multifactor models. It will increase the efficiency of the model building process and their subsequent use for searching both short-term and long-term forecasts. In order to eliminate the effect of random components of the time series with an uneven distribution of the values of the electric consumption on the training process of the neural network as a non-linear forecasting model, we suggest its preliminary preparation using the Kalman filter. Further optimization of the neural network topology is carried out on the basis of a genetic algorithm that allows, at the mutation stage, to adaptively choose the type of structure transformation most suitable for a given network configuration. Ref. 24, pic. 3

A System Dynamics Modeling Support System Based on Computational Intelligence

System dynamics (SD) is a complex systems modeling and simulation approach with wide ranging applications in various science and engineering disciplines. While subject matter experts lead most of the model building, recent advances have attempted to bring system dynamics closer to fast growing fields such as data sciences. This may prove promising for the development of novel support methods that augment human cognition and improve efficiencies in the model building process. A few different directions have been explored recently to support individual modeling stages, such as the generation of model structure, model calibration and policy optimization. However, an integrated approach that supports across the board modeling process is still missing. In this paper, a prototype integrated modeling support system is presented for the purpose of supporting the modelers at each stage of the process. The proposed support system facilitates data-driven inferring of causal loop diagrams (CLDs), stock-flow diagrams (SFDs), model equations and the estimation of model parameters using computational intelligence (CI) techniques. The ultimate goal of the proposed system is to support the construction of complex models, where the human power is not enough. With this goal in mind, we demonstrate the working and utility of the proposed support system. We have used two well-known synthetic reality case studies with small models from the system dynamics literature, in order to verify the support system performance. The experimental results showed the effectiveness of the proposed support system to infer close model structures to target models directly from system time-series observations. Future work will focus on improving the support system so that it can generate complex models on a large scale.

Payload Based Internet Worm Disclosure using Neural Network

With the capacity of contaminating a huge number of hosts, worms speak to a noteworthy danger to the Internet. The identification against Internet worms is generally an open issue. Web worms represent a genuine danger to PC security. Conventional methodologies utilizing marks to identify worms posture little risk to the zero day assaults. The focal point of this exploration is moving from utilizing mark examples to distinguishing the vindictive conduct showed by the Internet worms. This paper displays an original thought of separating stream level highlights that can distinguish worms from clean projects utilizing information mining method, for example, neural system classifier. Our approach demonstrated 97.90% recognition rate on Internet worms whose information was not utilized as a part of the model building process

Multi-Dimensional Penalized Hazard Model with Continuous Covariates: Applications for Studying Trends and Social Inequalities in Cancer Survival

SummaryDescribing the dynamics of patient mortality hazard is a major concern for cancer epidemiologists. In addition to time and age, other continuous covariates have often to be included in the model. For example, survival trend analyses and socio-economic studies deal respectively with the year of diagnosis and a deprivation index. Taking advantage of a recent theoretical framework for general smooth models, the paper proposes a penalized approach to hazard and excess hazard models in time-to-event analyses. The baseline hazard and the functional forms of the covariates were specified by using penalized natural cubic regression splines with associated quadratic penalties. Interactions between continuous covariates and time-dependent effects were dealt with by forming a tensor product smooth. The smoothing parameters were estimated by optimizing either the Laplace approximate marginal likelihood criterion or the likelihood cross-validation criterion. The regression parameters were estimated by direct maximization of the penalized likelihood of the survival model, which avoids data augmentation and the Poisson likelihood approach. The implementation proposed was evaluated on simulations and applied to real data. It was found to be numerically stable, efficient and useful for choosing the appropriate degree of complexity in overall survival and net survival contexts; moreover, it simplified the model building process.

Bayesian Network Modeling in Discovering Risk Factors of Dental Caries in Three-Year-Old Children

<br/><b>Purpose</b> - The aim of this study was to use probabilistic graphical models to determine dental caries risk factors in three-year-old children. The analysis was conducted on the basis of the questionnaire data and resulted in building probabilistic graphical models to investigate dependencies among the features gathered in the surveys on dental caries. <br/><b>Materials and Methods</b> - The data available in this analysis came from dental examinations conducted in children and from a questionnaire survey of their parents or guardians. The data represented 255 children aged between 36 and 48 months. Self-administered questionnaires contained 34 questions of socioeconomic and medical nature such as nutritional habits, wealth, or the level of education. The data included also the results of oral examination by a dentist. We applied the Bayesian network modeling to construct a model by learning it from the collected data. The process of Bayesian network model building was assisted by a dental expert. <br/><b>Results</b> - The model allows to identify probabilistic relationships among the variables and to indicate the most significant risk factors of dental caries in three-year-old children. The Bayesian network model analysis illustrates that cleaning teeth and falling asleep with a bottle are the most significant risk factors of dental caries development in three-year-old children, whereas socioeconomic factors have no significant impact on the condition of teeth. <br/><b>Conclusions</b> - Our analysis results suggest that dietary and oral hygiene habits have the most significant impact on the occurrence of dental caries in three-year-olds.

COMPARING MODEL-BUILDING PROCESS: A MODEL PROSPECTIVE TEACHERS USED IN INTERPRETING STUDENTS’ MATHEMATICAL THINKING

Mathematical thinking is an important aspect of mathematics education and, therefore, also needs to be understood by prospective teachers. Prospective teachers should have the ability to analyze and interpret students’ mathematical thinking. Comparing model is one of the interpretation models from Wilson, Lee, and Hollebrands. This article will describe the prospective teacher used the model of the building process in interpretation students' mathematical thinking. Subjects selected by considering them in following the students’ strategies in solving the Building Construction Problem. Comparing model is a model of interpretation in which a person interprets student thinking based on student work. There are two types comparing model building process prospective teacher use in interpreting students’ mathematical thinking ie. comparing work and comparing knowledge. In comparing works, prospective teachers use an external representation rubric. This is used to analyze student activities in order to provide an interpretation that is comparing the work of students with their own work. In comparing knowledge, prospective teachers use internal representation rubrics to provide interpretation by comparing the students' work with their knowledge or thought.

Dynamic Autoselection and Autotuning of Machine Learning Models for Cloud Network Analytics

Cloud network monitoring data is dynamic and distributed. Signals to monitor the cloud can appear, disappear or change their importance and clarity over time. Machine learning (ML) models tuned to a given data set can therefore quickly become inadequate. A model might be highly accurate at one point in time but may lose its accuracy at a later time due to changes in input data and their features. Distributed learning with dynamic model selection is therefore often required. Under such selection, poorly performing models (although aggressively tuned for the prior data) are retired or put on standby while new or standby models are brought in. The well-known method of Ensemble ML (EML) may potentially be applied to improve the overall accuracy of a family of ML models. Unfortunately, EML has several disadvantages, including the need for continuous training, excessive computational resources, requirement for large training datasets, high risks of overfitting, and a time-consuming model-building process. In this paper, we propose a novel cloud methodology for automatic ML model selection and tuning that automates model building and selection and is competitive with existing methods. We use unsupervised learning to better explore the data space before the generation of targeted supervised learning models in an automated fashion. In particular, we create a Cloud DevOps architecture for autotuning and selection based on container orchestration and messaging between containers, and take advantage of a new autoscaling method to dynamically create and evaluate instantiations of ML algorithms. The proposed methodology and tool are demonstrated on cloud network security datasets.

Philosophy of Economics: The Constructivist and Scientific Realist Interpretation of Macroeconomics

The contemporary methodological debate focuses on studying the relation between models and reality regarding idealization and resemblance. Limited attention is put on the process of model building. The study focuses its attention on the latter. The purpose of this article is twofold. First, we reconstruct the philosophical presuppositions of the participants of a cliometric debate. Second, we employ the Fleckian constructivist philosophy with the aim of offering an account of how econometric facts are constructed. Contrary to our reconstruction of the process of ‘observing’ macroeconomic phenomena, econometricians ground their viewpoint in the scientific-realist and neopositivist philosophies of science.

Influence of the range of data on the performance of ANN- and SVM- based time series models for reproducing groundwater level observations

In the present study, we designed time series models for predicting groundwater level fluctuations using an artificial neural network (ANN) and a support vector machine (SVM). To estimate the model sensitivity to the range of data set for the model building, numerical tests were conducted using hourly measured groundwater level data at a coastal aquifer of Jeju Island in South Korea. The model performance of the two models is similar and acceptable when the range of input variable lies within the data set for the model building. However, when the range of input variables is beyond it, both the models showed abnormal prediction results: an oscillation for the ANN model and a constant value for SVM. The result of the numerical tests indicates that it is necessary to obtain various types of input and output variables and assign them to the model building process for the success of design time series models of groundwater level prediction.

Distribution models of estuarine fish species: The effect of sampling bias, species ecology and threshold selection on models' accuracy

Species distribution models (SDMs) relate presence/absence data to environmental variables, allowing to predict species environmental requirements and potential distribution. They have been increasingly used in fields such as ecology, biogeography and evolution, and often support conservation priorities and strategies. Thus, it becomes crucial to understand how trustworthy and reliable their predictions are. Different approaches, such as using ensemble methods (combining forecasts of different single models), or applying the most suitable threshold to transform continuous probability maps into species presences or absences, have been used to reduce model-based uncertainty. Taking into account the influence of biased sampling imprecision in species location, small datasets and species ecological characteristics, may also help to detect and compensate for uncertainty in the model building process. To investigate the effect of applying an ensemble approach, several threshold selection criteria and different datasets representing seasonal and spatial sampling bias, on models' accuracy, SDMs were built for four estuarine fish species with distinct use of the estuarine systems. Overall, predictions obtained with the ensemble approach were more accurate. Variability in accuracy metrics obtained with the nine threshold selection criteria applied was more pronounced for species with low prevalence and when sensitivity was calculated. Higher values of accuracy measures were registered with the threshold that maximizes the sum of sensitivity and specificity, and the threshold where the predicted prevalence equals the observed, whereas the 0.5 cut-off was unreliable, originating the lowest values for these metrics. Accuracy of models created from a spatially biased sampling was overall higher than accuracy of models created with a seasonally biased sampling or with the multi-year database created and this pattern was consistently obtained for marine migrant species, which use estuaries as nursery areas, presenting a seasonally and regular use of these ecosystems. The ecological dependence between these fish species and estuaries may add difficulties in the model building process, and needs to be taken into account, to improve their accuracy. The present study highlights the need for a thorough analysis of the critical underlying issues of the complete model building process to predict the distribution of estuarine fish species, due to the particular and dynamic nature of these ecosystems.

Examining Pathways between Bully Victimization, Depression, & School Belonging Among Early Adolescents

The relationship between bully victimization and depression has been examined extensively with prior research showing long-term cascade of problems stemming from both exposure to victimization and depressive symptomology. However, prior research has failed to consider how protective factors may mitigate these long-term problems. Three theoretical models were tested: the interpersonal risk model, symptom driven model, and transactional model. The present study employs a novel statistical technique to explore longitudinal reciprocal associations among bullying, depression, and school belonging in a sample of 2177 middle school students (ages 11 to 15) in a Midwestern state. We used a model building process to explore the overall association between bully victimization, depression, and school belonging as well as a multi-group model in which models were estimated for boys and girls, separately. In our overall model, results indicated support for both symptom driven and interpersonal risk models. However, we did not find any significant buffering effect of school belonging. In our multi-group model, we found support for a buffering effect of school belonging for girls, but not boys. School belonging buffered long term problems associated with experiences of bully victimization via reductions in depression. Our findings point to the broader concept of school structure being differentially supportive and protective for various demographic groups and the need to consider the entire social ecology of a school when planning and implementing prevention interventions.

Visualization in Bayesian Workflow

AbstractBayesian data analysis is about more than just computing a posterior distribution, and Bayesian visualization is about more than trace plots of Markov chains. Practical Bayesian data analysis, like all data analysis, is an iterative process of model building, inference, model checking and evaluation, and model expansion. Visualization is helpful in each of these stages of the Bayesian workflow and it is indispensable when drawing inferences from the types of modern, high dimensional models that are used by applied researchers.

Methods of Computational Analysis in Kidney Development.

This chapter reviews some currently available methodologies for constructing mathematical models in kidney development. Mammalian nephrogenesis is a complex biological process, which in its earliest stages involves migration, condensation, proliferation, and differentiation of metanephric mesenchymal (MM) cells interacting with the uroepithelial cells of the ureteric bud (UB). First, the mathematical modelling in biology is generally described. Secondly, some accounts to biological pattern formation in modelling are given in general, including models that transcend the Turing model. This is followed by a short assessment on the main branch of models in the kidney development, the evaluation of the branching morphogenesis of the kidney. Finally, two alternative models in the early kidney development processes are given as an example. They also elucidate the difficulties in the model building process. Firstly, a computational model building with the CompuCell3D program for the early nephron progenitor cell movements with the key extracellular signaling effectors is depicted. This collective migration leads to the first pretubular aggregate (PTA). The simulation parameters of the program imitate the program's cell sorting example with different adhesions and chemoattractants. The program utilizes Cellular Potts Model (CPM) to describe the development. Secondly, an example of PTA to renal vesicle (RV) transition modelling is described. In that case, the model is unique, where the model process is based on the chemoattractants from UB.

Repair Process Models Containing Non-Free-Choice Structures Based on Logic Petri Nets

Process mining is an important technique for complex operational process management. It aims to extract valuable process-related information and build process models from event logs of enterprise information systems. Process models can describe the characteristics and verify the correctness of the system with the function of providing feedback to system designers. However, there may be some deviations between the activities in event logs and a process model because of the system upgrade and business process improvement. Thus the original model needs to be repaired because it does not describe the real business process well. For the models with non-free-choice structures, event logs can be replayed by the existing repair methods, but their repaired models are quite different from the business processes. Therefore, a new model repair method is proposed based on logic Petri nets in this paper. First, the concept of transition pairs and successor relationship is proposed. A successor relationship matrix is constructed. Then, the position to be repaired can be determined by traversing the transition pairs. Finally, the correctness and effectiveness of the proposed method are illustrated in our experiments.

Try, try again: Lessons learned from success and failure in participatory modeling

Participatory Modeling (PM) is becoming increasingly common in environmental planning and conservation, due in part to advances in cyberinfrastructure as well as to greater recognition of the importance of engaging a diverse array of stakeholders in decision making. We provide lessons learned, based on over 200 years of the authors’ cumulative and diverse experience, about PM processes. These include successful and, perhaps more importantly, not-so-successful trials. Our collective interdisciplinary background has supported the development, testing, and evaluation of a rich range of collaborative modeling approaches. We share here what we have learned as a community of participatory modelers, within three categories of reflection: a) lessons learned about participatory modelers; b) lessons learned about the context of collaboration; and c) lessons learned about the PM process. First, successful PM teams encompass a variety of skills beyond modeling expertise. Skills include: effective relationship-building, openness to learn from local experts, awareness of personal motivations and biases, and ability to translate discussions into models and to assess success. Second, the context for collaboration necessitates a culturally appropriate process for knowledge generation and use, for involvement of community co-leads, and for understanding group power dynamics that might influence how people from different backgrounds interact. Finally, knowing when to use PM and when not to, managing expectations, and effectively and equitably addressing conflicts is essential. Managing the participation process in PM is as important as managing the model building process. We recommend that PM teams consider what skills are present within a team, while ensuring inclusive creative space for collaborative exploration and learning supported by simple yet relevant models. With a realistic view of what it entails, PM can be a powerful approach that builds collective knowledge and social capital, thus helping communities to take charge of their future and address complex social and environmental problems.