Review
Thermodynamics
Tingting Li, Yang Zhao, Ke Yan, Kai Zhou, Chaobo Zhang, Xuejun Zhang
Summary: Probabilistic graphical models are effective in addressing various issues in energy systems, with static models handling incomplete or uncertain information and dynamic models accurately predicting energy consumption, occupancy, and failures. A unified framework combining knowledge-driven and data-driven PGMs is suggested for better performance, with the need for universal PGM-based approaches adaptable to different energy systems and hybrid algorithms integrating advanced techniques for improved results.
BUILDING SIMULATION
(2022)
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, Information Systems
Junji Jiang, Likang Wu, Hongke Zhao, Hengshu Zhu, Wei Zhang
Summary: Stock movement forecasting is often treated as a sequence prediction task using time series data. While deep learning models have been increasingly employed for fitting dynamic stock time series, few of them have focused on understanding the internal dynamics of the market system. To address this, the proposed HMM-ALSTM framework integrates the Hidden Markov Model (HMM) into the deep learning process, allowing for the discovery of hidden states and patterns that contribute to the stock time series data.
INFORMATION PROCESSING & MANAGEMENT
(2023)
Article
Economics
Zijian Zeng, Meng Li
Summary: A Bayesian median autoregressive model was developed for time series forecasting, utilizing time-varying quantile regression at the median and a Laplace error instead of Gaussian error. Model parameters were estimated using Markov chain Monte Carlo, with Bayesian model averaging and model selection used to address model uncertainty. The methods showed favorable predictive performance in real data applications.
INTERNATIONAL JOURNAL OF FORECASTING
(2021)
Article
Chemistry, Analytical
Ting Lin, Miao Wang, Min Yang, Xu Yang
Summary: This paper addresses the issues with commonly used methods in mining time series data by proposing a novel approach that utilizes Wasserstein distance and autoencoder to learn discrete features and hidden Markov model to learn continuous features. The two models are then stacked to create an ensemble model with lower computational complexity and comparable classification accuracy to state-of-the-art methods.
Article
Computer Science, Artificial Intelligence
Jinbo Li, Witold Pedrycz, Xianmin Wang, Peng Liu
Summary: This study introduces a fuzzy model based on Hidden Markov Model (HMM) for time series prediction. The model uses fuzzy rules to describe the relationship between input and output time series and employs HMM to capture the temporal behavior of multivariate time series. Experimental results show that the proposed model outperforms fuzzy rule-based models without involving HMMs.
Article
Mathematics
Byeongheon Lee, Joowon Park, Yongku Kim
Summary: A hidden Markov model (HMM) is a useful tool for modeling dependent heterogeneous phenomena and finding factors that affect real-world events. It differs from traditional methods by using state variables and mixture distributions to model hidden states. HMMs can be extended to include covariates and are applied in this study to find factors affecting the growth of matsutake mushrooms.
Article
Computer Science, Artificial Intelligence
Qin Tao, Yajing Si, Fali Li, Peiyang Li, Yuqin Li, Shu Zhang, Feng Wan, Dezhong Yao, Peng Xu
Summary: This study revealed the dynamic process of decision response and feedback in gambling, showing that different decisions and feedback lead to distinct brain network activities.
INTERNATIONAL JOURNAL OF NEURAL SYSTEMS
(2021)
Article
Computer Science, Interdisciplinary Applications
Jan G. De Gooijer, Gustav Eje Henter, Ao Yuan
Summary: This article introduces a conditional density estimation method, called KCDE-HMM, which combines kernel-based estimation and hidden Markov models. The method provides conditional density estimates for time series data and demonstrates superior predictive performance compared to benchmark methods, particularly for moderate-to-large sample sizes.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2022)
Article
Neurosciences
Yuhua Yu, Yongtaek Oh, John Kounios, Mark Beeman
Summary: This study uses unsupervised machine learning to analyze EEG data and reveals different cognitive processes involved in problem-solving, particularly dynamic features related to insight and analysis.
Article
Computer Science, Interdisciplinary Applications
Federico Castelletti, Stefano Peluso
Summary: The paper introduces a Bayesian methodology for structure learning of categorical essential graphs, deriving a closed-form expression for the marginal likelihood of a categorical essential graph model. The methodology combines constructive parameter prior elicitation with graph-driven likelihood decomposition, and is evaluated on simulated scenarios and real data with appreciable performance in comparison with state-of-the-art methods.
COMPUTATIONAL STATISTICS & DATA ANALYSIS
(2021)
Article
Computer Science, Theory & Methods
Yu Luo, David A. Stephens
Summary: This study focuses on modeling data generated by a latent continuous-time Markov jump process, proposing a reversible jump Markov chain Monte Carlo algorithm and applying it to model-based clustering analysis.
STATISTICS AND COMPUTING
(2021)
Article
Mathematics
Tao Li, Jinwen Ma
Summary: The mixture of Gaussian process functional regressions (GPFRs) assumes independent random processes with different temporal structures for generating time series or sample curves. However, in reality, these structures are randomly transferred from a long time scale. To address this limitation, we propose the hidden-Markov-based GPFR mixture model (HM-GPFR) that describes the curves using both fine- and coarse-level temporal structures. The model combines the Gaussian process at the fine level and the hidden Markov process at the coarse level, resulting in a random process with state switching dynamics. Additionally, a Bayesian-hidden-Markov-based GPFR mixture model (BHM-GPFR) is developed to enhance the robustness of the model by incorporating priori parameters. Experimental results show high prediction accuracy and interpretability.
Article
Multidisciplinary Sciences
Ha Yoon Song, Jae Ho Lee
Summary: With the advancement of geopositioning systems and mobile devices, much research is being conducted on geopositioning data. Map matching, a core preprocessing technique for trajectory data, is gaining attention, particularly the use of Hidden Markov Model (HMM) for map matching. However, the HMM model simplifies the dependency of time series data excessively, leading to incorrect matching results. In this research, a new algorithm called trendHMM map matching is proposed, which improves upon the assumptions of HMM by considering a wider range of dependencies and incorporating neighboring data.
Article
Engineering, Environmental
Guang Yang, Shenghui Fang, Wenbing Gong, Yaolong Zhao, Mengyu Ge
Summary: Time series land cover maps are important materials, but illogical transitions exist between different time phases. Limited ground truth evaluation cannot guide users well. A method based on joint probability is proposed to evaluate the reliability of time series land cover products.
STOCHASTIC ENVIRONMENTAL RESEARCH AND RISK ASSESSMENT
(2021)
Article
Computer Science, Theory & Methods
Yi-An Ma, Emily B. Fox, Tianqi Chen, Lei Wu
STATISTICS AND COMPUTING
(2019)
Article
Computer Science, Theory & Methods
Jack Baker, Paul Fearnhead, Emily B. Fox, Christopher Nemeth
STATISTICS AND COMPUTING
(2019)
Review
Mathematics, Interdisciplinary Applications
Ali Shojaie, Emily B. Fox
Summary: Granger causality, introduced over half a century ago, has become a popular tool for analyzing time series data in various fields. However, its validity in inferring causal relationships among time series has been continuously debated. Recent advances have addressed the limitations of earlier approaches, such as models for high-dimensional time series and accounting for nonlinearity and non-Gaussian observations.
ANNUAL REVIEW OF STATISTICS AND ITS APPLICATION
(2022)
Meeting Abstract
Endocrinology & Metabolism
Johannes Ferstad, Priya Prahalad, David M. Maahs, Emily Fox, Ramesh Johari, David Scheinker
Article
Public, Environmental & Occupational Health
Andrew C. Miller, Lauren A. Hannah, Joseph Futoma, Nicholas J. Foti, Emily B. Fox, Alexander D'Amour, Mark Sandler, Rif A. Saurous, Joseph A. Lewnard
Summary: Accurate measurement of daily infection incidence is crucial to epidemic response. This study introduces the robust incidence deconvolution estimator, which incorporates a regularization scheme to address stochastic delays. Comparison with existing estimators in simulation and real data analysis demonstrates its accuracy and stability.
Article
Endocrinology & Metabolism
Dessi P. Zaharieva, Ransalu Senanayake, Conner Brown, Brendan Watkins, Glenn Loving, Priya Prahalad, Johannes O. Ferstad, Carlos Guestrin, Emily B. Fox, David M. Maahs, David Scheinker
Summary: Stanford University has utilized algorithm-enabled patient prioritization and remote patient monitoring to improve clinical workflows and glucose control in youth with type 1 diabetes. The care model integrates continuous glucose monitoring data and aims to incorporate exercise data to better manage patients' needs and help healthcare professionals make informed decisions. Regular exercise is crucial for cardiovascular fitness and overall well-being, but can impact blood glucose levels. By integrating physical activity metrics, the model aims to identify whether patients are meeting exercise guidelines and provide clinically relevant information.
FRONTIERS IN ENDOCRINOLOGY
(2023)
Article
Clinical Neurology
Nina J. Ghosn, Kevin Xie, Akash R. Pattnaik, James J. Gugger, Colin A. Ellis, Elizabeth Sweeney, Emily Fox, John M. Bernabei, Jenaye Johnson, Jacqueline Boccanfuso, Brian Litt, Erin C. Conrad
Summary: Evaluating patients with drug-resistant epilepsy often requires inducing seizures by tapering antiseizure medications (ASMs) in the epilepsy monitoring unit (EMU). The relationship between ASM taper strategy, seizure timing, and severity remains unclear. In this study, we developed and validated a pharmacokinetic model of total ASM load and tested its association with seizure occurrence and severity in the EMU.
Article
Medicine, Research & Experimental
Bryan J. Bunning, Haley Hedlin, Jonathan H. Chen, Jody D. Ciolino, Johannes Opsahl Ferstad, Emily Fox, Ariadna Garcia, Alan Go, Ramesh Johari, Justin Lee, David M. Maahs, Kenneth W. Mahaffey, Krista Opsahl-Ong, Marco Perez, Kaylin Rochford, David Scheinker, Heidi Spratt, Mintu P. Turakhia, Manisha Desai
Summary: Clinical trials are incorporating real-world data sources to enhance generalizability of findings and overall trial scale and efficiency. However, the informatic complexity of the data requires a robust data science infrastructure, and monitoring data and safety must evolve to protect the rigor of clinical trials.
JOURNAL OF CLINICAL AND TRANSLATIONAL SCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Alex Tank, Ian Covert, Nicholas Foti, Ali Shojaie, Emily B. Fox
Summary: By using structured MLPs and RNNs with sparsity-inducing penalties on weights, we can extract the Granger causal structure and efficiently capture long-range dependencies. Our neural Granger causality methods outperform state-of-the-art nonlinear methods on challenging datasets, demonstrating the utility of deep learning beyond large dataset prediction scenarios.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2022)
Meeting Abstract
Cardiac & Cardiovascular Systems
Ransalu Senanayake, Johannes O. Ferstad, Isha Thapa, Flavia Giammarino, Megana Vasu, Dessi Zaharieva, Priya Prahalad, David M. Maahs, David N. Rosenthal, Fatima Rodriguez, Nicholas Bambos, Daniel Miller, Andrew Shin, Stephen J. Roth, Carlos Guestrin, Emily B. Fox, David Scheinker
Article
Mathematics, Applied
Alex Tank, Xiudi Li, Emily B. Fox, Ali Shojaie
Summary: The study proposes a framework for learning Granger causality networks for multivariate categorical time series based on the MTD model. By recasting inference in the MTD as a convex problem, the study addresses the issues of nonconvex objective and nonidentifiability, and demonstrates consistency in high dimensions. Additionally, the study introduces the mLTD model as a baseline for comparison.
SIAM JOURNAL ON MATHEMATICS OF DATA SCIENCE
(2021)
Proceedings Paper
Automation & Control Systems
Christopher Xie, Emily Fox, Zaid Harchaoui
2019 INTERNATIONAL CONFERENCE ON ROBOTICS AND AUTOMATION (ICRA)
(2019)
Article
Computer Science, Interdisciplinary Applications
Jack Baker, Paul Fearnhead, Emily B. Fox, Christopher Nemeth
JOURNAL OF STATISTICAL SOFTWARE
(2019)
Article
Automation & Control Systems
Joelle Pineau, Philippe Vincent-Lamarre, Koustuv Sinha, Vincent Lariviere, Alina Beygelzimer, Florence d'Alche-Buc, Emily Fox, Hugo Larochelle
Summary: Ensuring the reliability of results in machine learning research presents a challenge. Reproducibility is a crucial step in verifying research findings and promoting open and accessible research in the scientific community.
JOURNAL OF MACHINE LEARNING RESEARCH
(2021)
Article
Mathematics, Applied
Christopher Aicher, Yi-An Ma, Nicholas J. Foti, Emily B. Fox
SIAM JOURNAL ON MATHEMATICS OF DATA SCIENCE
(2019)
Article
Computer Science, Artificial Intelligence
Franz Baader, Oliver Fernandez Gil
Summary: This article introduces the problem of using traditional logic-based knowledge representation languages in AI applications, which can lead to complex definitions and difficult reasoning. To overcome this, the article proposes a new concept constructor and utilizes graded membership functions to define the semantics. Additionally, a class of concept measures is introduced and the algorithmic properties of the corresponding logic are analyzed.
ARTIFICIAL INTELLIGENCE
(2024)
Article
Computer Science, Artificial Intelligence
Chenjia Bai, Lingxiao Wang, Jianye Hao, Zhuoran Yang, Bin Zhao, Zhen Wang, Xuelong Li
Summary: Offline Reinforcement Learning has shown promising results in learning task-specific policies. However, directly sharing datasets from other tasks exacerbates the distribution shift issue in offline RL. In this paper, we propose an uncertainty-based Multi-Task Data Sharing approach that provides a unified framework for offline RL and resolves the distribution shift problem. The experimental results show that our algorithm outperforms previous state-of-the-art methods in challenging MTDS problems.
ARTIFICIAL INTELLIGENCE
(2024)
Article
Computer Science, Artificial Intelligence
Nemanja Ilic, Dejan Dasic, Miljan Vucetic, Aleksej Makarov, Ranko Petrovic
Summary: This paper proposes a novel adaptive consensus-based learning algorithm for automated and distributed web hacking. The algorithm aims to assist ethical hackers in conducting legitimate penetration testing and improving web security by efficiently identifying system vulnerabilities at an early stage. Through the use of interconnected intelligent agents and a decentralized communication scheme, the algorithm achieves fast convergence and outperforms existing schemes in terms of scalability and hacking times. This algorithm provides valuable insights and opportunities for system security administrators to effectively address identified shortcomings and vulnerabilities.
ARTIFICIAL INTELLIGENCE
(2024)
Article
Computer Science, Artificial Intelligence
Yusuke Kawamoto, Tetsuya Sato, Kohei Suenaga
Summary: This paper proposes a new approach for formally describing the requirement for statistical inference and checking the appropriate use of statistical methods in programs. The authors define a belief Hoare logic (BHL) for formalizing and reasoning about statistical beliefs acquired through hypothesis testing. Examples demonstrate the usefulness of BHL in reasoning about practical issues in hypothesis testing, while also discussing the importance of prior beliefs in acquiring statistical beliefs.
ARTIFICIAL INTELLIGENCE
(2024)
