Inference in Hidden Markov Models with Explicit State Duration Distributions

Dewar, M. E.; Wiggins, Chris H.; Wood, Frank

doi:10.1109/lsp.2012.2184795

Cited by 35 publications

(20 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The second difference between our model and an ordinary HMM is that, when an individual enters a new state Z j , they also draw a number of timesteps—a duration —to remain in that state before transitioning to the next state, an idea from hidden semi-Markov models [ 65 ] (HSMMs) and explicit duration hidden Markov models [ 21 , 37 ] (EDHMMs). In ordinary HMMs, the time spent in a state is drawn from a geometric distribution [ 17 ], but as we discuss below, this parameterization is a poor approximation when modeling many real-world cycles; drawing state durations from arbitrary duration distributions allows for more flexible and realistic cycle modeling. Figure 1 illustrates a path an individual might take through hidden states over the course of a single cycle.…”

Section: Proposed Modelmentioning

confidence: 99%

Modeling Individual Cyclic Variation in Human Behavior

Pierson

Althoff

Leskovec

2018

Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW '18

View full text Add to dashboard Cite

Cycles are fundamental to human health and behavior. Examples include mood cycles, circadian rhythms, and the menstrual cycle. However, modeling cycles in time series data is challenging because in most cases the cycles are not labeled or directly observed and need to be inferred from multidimensional measurements taken over time. Here, we present Cyclic Hidden Markov Models (CyH-MMs) for detecting and modeling cycles in a collection of multidimensional heterogeneous time series data. In contrast to previous cycle modeling methods, CyHMMs deal with a number of challenges encountered in modeling real-world cycles: they can model multivariate data with both discrete and continuous dimensions; they explicitly model and are robust to missing data; and they can share information across individuals to accommodate variation both within and between individual time series. Experiments on synthetic and real-world health-tracking data demonstrate that CyHMMs infer cycle lengths more accurately than existing methods, with 58% lower error on simulated data and 63% lower error on real-world data compared to the best-performing baseline. CyHMMs can also perform functions which baselines cannot: they can model the progression of individual features/symptoms over the course of the cycle, identify the most variable features, and cluster individual time series into groups with distinct characteristics. Applying CyHMMs to two real-world health-tracking datasets—of human menstrual cycle symptoms and physical activity tracking data—yields important insights including which symptoms to expect at each point during the cycle. We also find that people fall into several groups with distinct cycle patterns, and that these groups differ along dimensions not provided to the model. For example, by modeling missing data in the menstrual cycles dataset, we are able to discover a medically relevant group of birth control users even though information on birth control is not given to the model.

show abstract

Section: Proposed Modelmentioning

confidence: 99%

Modeling Individual Cyclic Variation in Human Behavior

Pierson

Althoff

Leskovec

2018

Proceedings of the 2018 World Wide Web Conference on World Wide Web - WWW '18

View full text Add to dashboard Cite

show abstract

“…However, modelling these correlations provides the promise of 'borrowing strength' across EUs to impute CPs, where data are scarce, as is commonly possible in Bayesian analysis [40]. The specification of infinite hidden Markov models, particularly allowing for variable dwelling times, may prove a fruitful extension [41]. The generation of parameters for the DM distribution may also be extended by modelling how more recent cultures arise from older ones according to, for instance, a hierarchical Dirichlet process, or by providing additional parameteric flexibility afforded by a logistic multivariate normal distribution [42].…”

Section: Discussionmentioning

confidence: 99%

Mixture model of pottery decorations from Lake Chad Basin archaeological sites reveals ancient segregation patterns

2016

View full text Add to dashboard Cite

We present a new statistical approach to analysing an extremely common archaeological data type-potsherds-that infers the structure of cultural relationships across a set of excavation units (EUs). This method, applied to data from a set of complex, culturally heterogeneous sites around the Mandara mountains in the Lake Chad Basin, helps elucidate cultural succession through the Neolithic and Iron Age. We show how the approach can be integrated with radiocarbon dates to provide detailed portraits of cultural dynamics and deposition patterns within single EUs. In this context, the analysis supports ancient cultural segregation analogous to historical ethnolinguistic patterning in the region. We conclude with a discussion of the many possible model extensions using other archaeological data types.

show abstract

“…in [41]. Since then it has been used in different contexts such as general modeling in case of explicit state duration HMMs [42] but also for speech synthesis [43] or the generation and decoding of temporal sequences, such as music [44].…”

Section: Length Modeling For Temporal Sequencesmentioning

confidence: 99%

A Hybrid RNN-HMM Approach for Weakly Supervised Temporal Action Segmentation

Kuehne

Richard

Gall

2020

IEEE Trans. Pattern Anal. Mach. Intell.

View full text Add to dashboard Cite

Action recognition has become a rapidly developing research field within the last decade. But with the increasing demand for large scale data, the need of hand annotated data for the training becomes more and more impractical. One way to avoid frame-based human annotation is the use of action order information to learn the respective action classes. In this context, we propose a hierarchical approach to address the problem of weakly supervised learning of human actions from ordered action labels by structuring recognition in a coarse-to-fine manner. Given a set of videos and an ordered list of the occurring actions, the task is to infer start and end frames of the related action classes within the video and to train the respective action classifiers without any need for hand labeled frame boundaries. We address this problem by combining a framewise RNN model with a coarse probabilistic inference. This combination allows for the temporal alignment of long sequences and thus, for an iterative training of both elements. While this system alone already generates good results, we show that the performance can be further improved by approximating the number of subactions to the characteristics of the different action classes as well as by the introduction of a regularizing length prior. The proposed system is evaluated on two benchmark datasets, the Breakfast and the Hollywood extended dataset, showing a competitive performance on various weak learning tasks such as temporal action segmentation and action alignment.

show abstract

Inference in Hidden Markov Models with Explicit State Duration Distributions

Cited by 35 publications

References 11 publications

Modeling Individual Cyclic Variation in Human Behavior

Modeling Individual Cyclic Variation in Human Behavior

Mixture model of pottery decorations from Lake Chad Basin archaeological sites reveals ancient segregation patterns

A Hybrid RNN-HMM Approach for Weakly Supervised Temporal Action Segmentation

Contact Info

Product

Resources

About