Article
Statistics & Probability
Roberto Colombi, Sabrina Giordano, Maria Kateri
Summary: This work analyzes longitudinal ordinal responses by proposing an approach that models the temporal dynamics of a latent trait of interest and the answering behaviors influenced by response styles through hidden Markov models (HMMs) with two latent components. The proposed model enables the modeling of independent latent traits and changes over time, as well as controlling for individual characteristics.
STATISTICAL METHODS AND APPLICATIONS
(2023)
Article
Computer Science, Theory & Methods
Salvatore D. Tomarchio, Antonio Punzo, Antonello Maruotti
Summary: This paper introduces HMMs for analyzing matrix-variate balanced longitudinal data, assuming a matrix normal distribution for each hidden state. The issue of potential overparameterization is addressed by using the eigen decomposition of covariance matrices, leading to 98 HMMs. An expectation-conditional maximization algorithm is discussed for parameter estimation. The proposed models are validated on simulated and real data sets.
STATISTICS AND COMPUTING
(2022)
Article
Statistics & Probability
Saiedeh Haji-Maghsoudi, Majid Sadeghifar, Ghodratollah Roshanaei, Hossein Mahjub
Summary: A multivariate hidden semi-Markov regression model was proposed to assess the effects of covariates on multiple responses in longitudinal data. Simulation studies were conducted to evaluate the properties of the model, and its application in hemodialysis patients' data was demonstrated.
COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
(2021)
Article
Psychology, Multidisciplinary
Mingrui Liang, Matthew D. Koslovsky, Emily T. Hebert, Darla E. Kendzor, Michael S. Businelle, Marina Vannucci
Summary: Continuous-time hidden Markov models are useful for studying the relationships between risk factors and outcomes, but variable selection methods have not been widely applied to this class of models. In this study, we develop a Bayesian continuous-time hidden Markov model with variable selection priors and provide an R package for practical application. The model is demonstrated on simulated data and applied to longitudinal data from a smoking cessation trial. The importance of adjusting for measurement error and biases in intensive longitudinal data is also discussed.
PSYCHOLOGICAL METHODS
(2023)
Article
Physics, Multidisciplinary
Ozgur Danisman, Umay Uzunoglu Kocer
Summary: Hidden Markov models are commonly used for modeling probabilistic structures with latent variables. They assume that observation symbols are conditionally independent and identically distributed, but in practice, this assumption may not always hold. The proposed model introduces a first-order Markov dependency between the current pair of hidden state-emitted observation symbol and the previous pair, which can better capture possible dependencies in real-life scenarios.
PHYSICA A-STATISTICAL MECHANICS AND ITS APPLICATIONS
(2021)
Article
Mathematics, Interdisciplinary Applications
Abdessatar Souissi, El Gheteb Soueidi
Summary: This paper aims to expand on previous research on quantum hidden Markov processes by introducing the concept of entangled hidden Markov processes. These are hidden Markov processes in which the hidden processes themselves are entangled Markov processes. The paper provides an explicit expression for the joint expectation of these processes and demonstrates that the approach also applies to the classical case.
CHAOS SOLITONS & FRACTALS
(2023)
Article
Ecology
Ron R. Togunov, Andrew E. Derocher, Nicholas J. Lunn, Marie Auger-Methe
Summary: Movement is the primary means by which animals obtain resources and avoid hazards, with most movement exhibiting directional bias related to environmental features. A new biased correlated random walk model is proposed to classify and characterize behaviors with directional bias, offering a more accurate method to analyze navigation and flight behaviors in animals.
METHODS IN ECOLOGY AND EVOLUTION
(2021)
Article
Mathematical & Computational Biology
Saiedeh Haji-Maghsoudi, Jan Bulla, Majid Sadeghifar, Ghodratollah Roshanaei, Hossein Mahjub
Summary: The combination of hidden Markov and semi-Markov models form a flexible modeling tool that can capture various latent patterns and dynamics in different types of data. These models have become increasingly popular in longitudinal studies due to their ability to handle different response types.
STATISTICS IN MEDICINE
(2021)
Article
Chemistry, Multidisciplinary
Norah Abanmi, Heba Kurdi, Mai Alzamel
Summary: The prevalence of malware attacks targeting IoT systems has raised concerns and emphasized the need for effective detection and defense mechanisms. However, detecting malware, especially those with new or unknown behaviors, is challenging. The main issue lies in its ability to hide, making it difficult to detect. Moreover, limited information on malware families restricts the availability of big data for analysis. This paper introduces a new Profile Hidden Markov Model (PHMM) for app analysis and classification in Android systems, while also dynamically identifying suspicious calls to reduce the risk of code execution. The experimental results demonstrate that the proposed Dynamic IoT malware Detection in Android Systems using PHMM (DIP) outperforms eight rival malware detection frameworks, achieving up to 96.3% accuracy at a 5% False Positive Rate (FP rate), 3% False Negative Rate (FN rate), and 94.9% F-measure.
APPLIED SCIENCES-BASEL
(2023)
Article
Computer Science, Artificial Intelligence
Zhengyi Xing, Yulong Qiao, Yue Zhao, Wenhui Liu
Summary: In this paper, a bag-of-models method based on the mixture of SHMMs is proposed to describe dynamic texture for dynamic texture classification. The effectiveness of this method is demonstrated on different benchmark datasets.
COMPUTER VISION AND IMAGE UNDERSTANDING
(2023)
Article
Computer Science, Artificial Intelligence
Hui Lan, Ziquan Liu, Janet H. Hsiao, Dan Yu, Antoni B. Chan
Summary: This article proposes a novel HMM-based clustering algorithm, which clusters HMMs through their densities and priors, and simultaneously learns posteriors for the novel HMM cluster centers that compactly represent the structure of each cluster. The numbers K and S are automatically determined in two ways.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Physics, Fluids & Plasmas
Harrison Hartle, Fragkiskos Papadopoulos, Dmitri Krioukov
Summary: This study introduces natural temporal extensions of static hidden-variable network models with stochastic dynamics of hidden variables and links, exploring the structural deviations and level of persistence compared to static models. By controlling the dynamic parameters of hidden variables and links, the equivalence between snapshots of networks in dynamic and static models under different conditions is examined. The authors also discuss qualitative resemblances between these dynamic network models and real systems, speculating on the presence of out-of-equilibrium links with respect to hidden variables in some real networks.
Article
Geochemistry & Geophysics
Zetao Wang, Gang Li, Le Yang
Summary: This letter introduces a novel method for dynamic hand gesture recognition based on micro-Doppler radar signatures. The method utilizes short-time Fourier transform to obtain time-frequency spectrograms and models them with a hidden Gauss-Markov model for recognition. Experimental results show strong generalization ability in radar gesture recognition, even in low SNR and unknown user scenarios.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2021)
Article
Biology
Lauren Hoskovec, Matthew D. Koslovsky, Kirsten Koehler, Nicholas Good, Jennifer L. Peel, John Volckens, Ander Wilson
Summary: This paper presents an infinite hidden Markov model for multiple asynchronous multivariate time series with missing data. The model excels in estimating hidden states and imputing missing data through beam sampling and Bayesian multiple imputation algorithm. The model performs well in simulation studies and real-case validation, showing improvements in estimation and imputation compared to existing approaches.
Article
Computer Science, Artificial Intelligence
Carlos Puerto-Santana, Pedro Larranaga, Concha Bielza
Summary: This article introduces asymmetric hidden Markov models with feature saliencies, which are capable of simultaneously determining relevant variables/features and probabilistic relationships between variables during their learning phase. Comparing with other approaches, the proposed models have better or equal fitness and provide further data insights.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)