Jean-Michel Tremblay · Digital Repository at the University of Maryland (University of Maryland College Park) · 2016
This dissertation proposes statistical methods to formulate, estimate and apply complex transportation models. Two main problems are part of the analyses conducted and presented in this dissertation. The first method solves an econometric problem and is concerned with the joint estimation of models that contain both discrete and continuous decision variables. The use of ordered models along with a regression is proposed and their effectiveness is evaluated with respect to unordered models. Procedure to calculate and optimize the log-likelihood functions of both discrete-continuous approaches are derived, and difficulties associated with the estimation of unordered models explained. Numerical approximation methods based on the Genz algortithm are implemented in order to solve the multidimensional integral associated with the unordered modeling structure. The problems deriving from the lack of smoothness of the probit model around the maximum of the log-likelihood function, which makes the optimization and the calculation of standard deviations very difficult, are carefully analyzed. A methodology to perform out-of-sample validation in the context of a joint model is proposed. Comprehensive numerical experiments have been conducted on both simulated and real data. In particular, the discrete-continuous models are estimated and applied to vehicle ownership and use models on data extracted from the 2009 National Household Travel Survey. The second part of this work offers a comprehensive statistical analysis of free-flow speed distribution; the method is applied to data collected on a sample of roads in Italy. A linear mixed model that includes speed quantiles in its predictors is estimated. Results show that there is no road effect in the analysis of free-flow speeds, which is particularly important for model transferability. A very general framework to predict random effects with few observations and incomplete access to model covariates is formulated and applied to predict the distribution of free-flow speed quantiles. The speed distribution of most road sections is successfully predicted; jack-knife estimates are calculated and used to explain why some sections are poorly predicted. Eventually, this work contributes to the literature in transportation modeling by proposing econometric model formulations for discrete-continuous variables, more efficient methods for the calculation of multivariate normal probabilities, and random effects models for free-flow speed estimation that takes into account the survey design. All methods are rigorously validated on both real and simulated data.
Cited by 0Open access
Trofimov A.G. , Shamiev M.O. · Zenodo (CERN European Organization for Nuclear Research) · 2025
This paper presents a comparative analysis of various data-driven approaches for Sardoba Dam settlement and failure prediction, including statistical, machine learning, deep learning, hybrid, and fuzzy-based models. The selected methods—ARIMA, Artificial Neural Networks (ANN), Long Short-Term Memory (LSTM), CNN–LSTM hybrid models, Adaptive Neuro-Fuzzy Inference System (ANFIS), Type-2 Fuzzy Logic, and ensemble/regression techniques such as Random Forest (RF) and Support Vector Regression (SVR)—are evaluated in terms of predictive accuracy, robustness, interpretability, and applicability to complex dam deformation scenarios. The study highlights the strengths and limitations of each approach, showing that while traditional statistical models like ARIMA capture linear temporal trends effectively, deep learning and hybrid models (ANN, LSTM, CNN–LSTM, ANN–ARIMA) provide superior performance in modeling nonlinear and time-dependent behaviors. Fuzzy-based systems, including ANFIS and Type-2 fuzzy logic, offer advantages in handling uncertainty and imprecise data. Ensemble methods such as RF and regression-based SVR provide reliable predictions under noisy or limited datasets. Through this analysis, the paper aims to identify the most effective modeling frameworks for accurate and early prediction of Sardoba Dam settlement and potential failure.
Cited by 1Open access
Trofimov A.G. , Shamiev M.O. · Zenodo (CERN European Organization for Nuclear Research) · 2025
This paper presents a comparative analysis of various data-driven approaches for Sardoba Dam settlement and failure prediction, including statistical, machine learning, deep learning, hybrid, and fuzzy-based models. The selected methods—ARIMA, Artificial Neural Networks (ANN), Long Short-Term Memory (LSTM), CNN–LSTM hybrid models, Adaptive Neuro-Fuzzy Inference System (ANFIS), Type-2 Fuzzy Logic, and ensemble/regression techniques such as Random Forest (RF) and Support Vector Regression (SVR)—are evaluated in terms of predictive accuracy, robustness, interpretability, and applicability to complex dam deformation scenarios. The study highlights the strengths and limitations of each approach, showing that while traditional statistical models like ARIMA capture linear temporal trends effectively, deep learning and hybrid models (ANN, LSTM, CNN–LSTM, ANN–ARIMA) provide superior performance in modeling nonlinear and time-dependent behaviors. Fuzzy-based systems, including ANFIS and Type-2 fuzzy logic, offer advantages in handling uncertainty and imprecise data. Ensemble methods such as RF and regression-based SVR provide reliable predictions under noisy or limited datasets. Through this analysis, the paper aims to identify the most effective modeling frameworks for accurate and early prediction of Sardoba Dam settlement and potential failure.
Cited by 0Open access
Ahmed Alghamdi · Communications on Applied Nonlinear Analysis · 2024
The growing sophistication of malware in exisiting environment poses tough demands on its detection methods of another kind. This paper introduces a framework for cyber security improvement, a means to solve strongly the probability issue of how to determine the behavior algorithmically in your system. This research is applied to malware detection systems based on methods (Support Vector Machines, Random Forest) and Neural Networks that are robust against deception. Statistical analysis of security threats and a review of current methods used for malware detection are given at its outset. The proposed framework, with its statistical framework analysis, looks for patterns and anomalies that indicate malice. It then integrates cryptographic techniques onto data transmission, computation processes; that the core of our system remains untouched by potential disturbances. The concept is further elaborated with open-source code from the field, Machine Learning (SVM, Random Forest, Neural Networks), As if with this variety of tools we can now detect and classify types of malware that generate irregular patterns but do not necessarily resemble known signatures at all. Such a hybrid security mechanism as described here, combining the best of statistics, cryptographic security, and machine learning, gives a precision of detection beyond compare to today's systems. The experiments show large gains in both detection accuracy and response speed, demonstrating this whole-element approach's potential to better safeguard cybersecurity in all manner of fields. It may be foreseen that this proposed line of research could well become a key new tool, capable of adaptation and expansion with the needs expanding around us.
Cited by 0Open access
LI Tianxi · 2025
This hands-on seminar introduces modern statistical methods for social network data, from network fundamentals to advanced statistical network models and privacy considerations. Through reproducible R demonstrations, you will learn to implement and diagnose models, design valid tests for relational data, and apply principled, publishable workflows for high-impact network research.
Cited by 0
Roxana Elena CRISTEA, Petre BREZEANU · SHILAP Revista de lepidopterología · 2022
A wide range of statistical techniques are used in building scoring models. Most of these statistical models are applicable to build an efficient credit scoring system that can be effectively used for predictive purposes. Techniques such as weight-of-evidence measurement, regression analysis, discriminant analysis, probit analysis, logistic regression, linear programming, Cox's proportional hazard model, support vector machines, decision trees, neural networks, k-nearest-neighbor , genetic algorithms, and genetic programming are all techniques widely used in building credit scoring models by credit analysts, researchers, lenders, and computer software developers and vendors. Advanced statistical methods vs. traditional statistical methods: advanced statistical techniques such as neural networks and genetic programming offer an alternative to conventional statistical techniques such as discriminant analysis, probit analysis and logistic regression. The point of using sophisticated techniques such as neural networks is their ability to model highly complex functions, and of course this is in contrast to traditional linear techniques such as linear regression and linear discriminant analysis.
Cited by 0Open access
David B. Flora · 2017
Statistical methods in modern research increasingly entail developing, estimating and testing models for data. Rather than rigid methods of data analysis, the need today is for more flexible methods for modelling data.
In this logical, easy-to-follow and exceptionally clear book, David Flora provides a comprehensive survey of the major statistical procedures currently used. His innovative model-based approach teaches you how to:
• Understand and choose the right statistical model to fit your data
• Match substantive theory and statistical models
• Apply statistical procedures hands-on, with example data analyses
• Develop and use graphs to understand data and fit models to data
• Work with statistical modeling principles using any software package
• Learn by applying, with input and output files for R, SAS, SPSS, and Mplus.
Statistical Methods for the Social and Behavioural Sciences: A Model Based Approach is the essential guide for those looking to extend their understanding of the principles of statistics, and begin using the right statistical modeling method for their own data. It is particularly suited to second or advanced courses in statistical methods across the social and behavioural sciences.
Cited by 27
Olivier Jaquet, Roberto Carniel · Statistics in Volcanology · 2018
Statistics in Volcanology is a comprehensive guide to modern statistical methods applied in volcanology written by today's leading authorities. The volume aims to show how the statistical analysis of complex volcanological data sets, including time series, and numerical models of volcanic processes can improve our ability to forecast volcanic eruptions. Specific topics include the use of expert elicitation and Bayesian methods in eruption forecasting, statistical models of temporal and spatial patterns of volcanic activity, analysis of time series in volcano seismology, probabilistic hazard assessment, and assessment of numerical models using robust statistical methods. Also provided are comprehensive overviews of volcanic phenomena, and a full glossary of both volcanological and statistical terms.Statistics in Volcanology is essential reading for advanced undergraduates, graduate students, and research scientists interested in this multidisciplinary field.
Cited by 23Open access
Costanza Bonadonna · Statistics in Volcanology · 2018
Statistics in Volcanology is a comprehensive guide to modern statistical methods applied in volcanology written by today's leading authorities. The volume aims to show how the statistical analysis of complex volcanological data sets, including time series, and numerical models of volcanic processes can improve our ability to forecast volcanic eruptions. Specific topics include the use of expert elicitation and Bayesian methods in eruption forecasting, statistical models of temporal and spatial patterns of volcanic activity, analysis of time series in volcano seismology, probabilistic hazard assessment, and assessment of numerical models using robust statistical methods. Also provided are comprehensive overviews of volcanic phenomena, and a full glossary of both volcanological and statistical terms.Statistics in Volcanology is essential reading for advanced undergraduates, graduate students, and research scientists interested in this multidisciplinary field.
Cited by 62Open access
Carpita, Maurizio, Piccolo, Domenico; SHAPE, University of Napoli Federico II, Ciavolino, Enrico; Dipartimento di Filosofia e Scienze Sociali, Università del Salento · Università del Salento · 2015
Statistical methods, models and applications that assume the existence of not directly observable variables are today extensively used in every research area. This Special Issue, inspired by the Workshop " Statistics with UNObservable Variables & Statistical models for HumAn Perception and Evaluation '' ( Sunova & Shape 2014 ), which was held at the Department of Economics and Business of the University of Brescia (Italy) as part of the activities of the DMS StatLab – Statistical Laboratory "Data Methods and Systems'" in October 21, 2014, highlights the opportunities offered by some of the most advanced statistical approaches based on unobservable variables.
Cited by 0Open access
Diego Restrepo-Leal, Maryam Carrillo-Reales · OSF Preprints (OSF Preprints) · 2025
This time series meta-analysis will focus on evaluating the comparative accuracy (measured by absolute RMSE and MAE errors, and normalized nRMSE and nMAE errors) of Advanced Methods (including AI/ML, Deep Learning, Ensemble, and Hybrid methods) versus Classical Methods (pure Physical, Statistical, and Naive models). The objective is to compare their performance in estimating wind profiles at onshore and offshore sites. The central purpose is to determine whether integrating advanced learning-based techniques (characteristic of advanced methods) provides the most robust and efficient solution for modeling the complexity and instability of wind flow introduced by the critical influence of coastal dynamics or trade winds (such as sea breeze flow, low-level jet streams, or variable atmospheric stability). This environmental dynamic introduces a high risk of systematic bias in Classical models, particularly in pure Physical and simple statistical models. The results will provide evidence-based guidance for selecting appropriate modeling methodologies for these wind regimes, helping to reduce uncertainty and financial risk in the planning and evaluation of new wind projects in these areas.
Cited by 0
Cannon, William · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 2017
MSPolygraph is designed to identify peptides from mass spectrometry data using advanced statistical methods and multiple models of peptide fragmentation. It can be executed in either serial or parallel computer architectures.
Cited by 0Open access
Cannon, William · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 2017
MSPolygraph is designed to identify peptides from mass spectrometry data using advanced statistical methods and multiple models of peptide fragmentation. It can be executed in either serial or parallel computer architectures.
Cited by 0Open access
Anne Ruiz‐Gazen, Nathalie Vialaneix · 2015
We greatly appreciate the opportunity to read and discuss the paper “Visualizing statistical models: removing the blindfold” [42]. The article manages to close the gap between statistics and visualization by combining advanced methods from both fields to visualize statistical models and methods. The article provides a very nice overview of visualization tools for statistical models, which can be used to 1) understand the model itself (i.e., what the model says about the data), 2) assess its relevance (i.e., if the model is accurate to describe the data, if the model has been well trained...) and 3) evaluate the variability of a family of models, use this information to select a model within a family, evaluate its robustness and combine several models in a relevant manner.
Cited by 0
David J. Gibson · 2014
Abstract This chapter moves forward from Chapter 7 (planning, choosing, and using statistics) and introduces some more advanced statistical methods that are of particular importance to plant population ecologists. The general goal of these methods is to quantify the spatiotemporal dynamics of plant populations. The basis for ecological modelling is described and advanced methods are described in four sections: first- and second-order spatial pattern analysis (including tessellation models); life table response experiments (LTREs), survivorship curves, and matrix models; cellular automata models, individual-based dynamic population models (e.g., SORTIE), and integral projection models (IPMs); and population viability analysis (PVA). Methods of spatial analysis are illustrated through use of a completely mapped plant dataset. Matrix models are illustrated through reanalysis of a published example. Recommended R packages for each method are provided.
Cited by 0
José Carlos Román-Herrera, Martín Jesús Rodríguez-Peces, Julio Garzón-Roca · Applied Sciences · 2023
A comparative methodology between advanced statistical tools and physical-based methods is carried out to ensure their reliability and objectivity for the evaluation of co-seismic landslide hazard maps. To do this, an inventory of landslides induced by the 2011 Lorca earthquake is used to highlight the usefulness of these methods to improve earthquake-induced landslide hazard analyses. Various statistical models, such as logistic regression, random forest, artificial neural network, and support vector machine, have been employed for co-seismic landslide susceptibility mapping. The results demonstrate that machine learning techniques using principal components (especially, artificial neural network and support vector machine) yield better results compared to other models. In particular, random forest shows poor results. Artificial neural network and support vector machine approaches are compared to the results of physical-based methods in the same area, suggesting that machine learning methods can provide better results for developing co-seismic landslide susceptibility maps. The application of different advanced statistical models shows the need for validation with an actual inventory of co-seismic landslides to ensure reliability and objectivity. In addition, statistical methods require a great amount of data. The results establish effective land planning and hazard management strategies in seismic areas to minimize the damage of future co-seismic landslides.
Cited by 9Open access
Jochen Voß · Medical Entomology and Zoology · 2013
A comprehensive introduction to sampling-based methods in statistical computing The use of computers in mathematics and statistics has opened up a wide range of techniques for studying otherwise intractable problems. Sampling-based simulation techniques are now an invaluable tool for exploring statistical models. This book gives a comprehensive introduction to the exciting area of sampling-based methods. An Introduction to Statistical Computing introduces the classical topics of random number generation and Monte Carlo methods. It also includes some advanced met
Cited by 22Open access
Desagulier, Guillaume · SPIRE - Sciences Po Institutional REpository · 2025
Statistical methods have been widely applied to linguistic data analysis and have evolved from basic frequency studies to advanced computational models. This entry shows the progression from foundational statistical concepts (including Zipf's law, probability distributions, lexical diversity measures, and vocabulary growth) to core analytical techniques such as collocation analysis, term frequency weighting, and n-gram modeling. It extends to contemporary computational approaches: vector space models, word embeddings, topic modeling, and machine learning classification methods. This development reflects the field's evolution from descriptive corpus linguistics toward predictive computational frameworks that integrate statistical theory with natural language processing applications.
Cited by 0Open access
Rosa María Prol-Ledesma · QRU Quaderns de Recerca en Urbanisme · 2024
The exploration of energy resources, in particular geothermal energy, has evolved towards the quantitative evaluation of information produced during reconnaissance and at advanced stages. The introduction of the concepts of geothermal plays and play-fairway have provided the parameters used in the new statistical integration models. Knowledge-based integration models have been displaced by datadriven models using statistical methods and machine learning in the exploration and evaluation process, thus increasing the efficiency of geothermal reservoir discovery.
Cited by 0Open access
Satish S Banait · Panamerican mathematical journal. · 2024
Improving numerical models through advanced data analysis in computational mathematics is a major step forward in many areas of science and engineering. The main goal of this study is to find ways to make accuracy, speed, and the ability to predict better by combining advanced data analysis techniques with computer simulation techniques. The study looks at how machine learning algorithms, statistical methods, and traditional computer methods can work together to solve hard, multidimensional problems. Algorithms for machine learning, like neural networks and support vector machines, are used to find trends and connections in very big datasets. These algorithms can give us information that other methods might miss. These new ideas make it easier to make models and programs that are more realistic. Statistical methods, like regression analysis and Bayesian reasoning, are used to measure error and make computer results more reliable. The goal of the study is to lower the cost of computing and raise the rate at which computer models converge by using these methods. A big part of this study is using these advanced methods to solve problems in the real world, like in material science, fluid dynamics, and climate models. Case studies show how combining computer simulations and data analysis can help make predictions that are more accurate and help people make better decisions. The study also looks at the problems that come with how hard these advanced methods are to compute and how much data they need, and it suggests ways to fix these problems. This study shows how mixing advanced data analysis with computer models in computational mathematics can change things. The results show how important it is to use methods from different fields to solve difficult science and engineering issues. This makes it possible for more accurate and useful models that can lead to new ideas and discoveries in many areas. This combination not only makes numerical models more powerful, but it also creates new study and use opportunities in computer mathematics.
Cited by 0Open access