Discovery of Causal Time Intervals

Li, Zhenhui; Zheng, Guopei; Agarwal, Amal; Xue, Lingzhou; Lauvaux, Thomas

doi:10.1137/1.9781611974973.90

Cited by 12 publications

(4 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Schölkopf [10] looked conversely into how causality can be used in ML, especially semi-supervised learning, to enhance robustness by leveraging cross-domain invariant causal mechanisms. Since then, most research in this field has been built on the assumption that the generation of cause data and the causal mechanism (P (ef f ect|cause)) are independent [30], [33]- [35], overlooking the temporality of causal relationships [44]. We approach causality from a unique perspective where time is equally critical and informative.…”

Section: Related Workmentioning

confidence: 99%

A Self-Labeling Method for Adaptive Machine Learning by Interactive Causality

Ren,

Yen,

2024

IEEE Trans. Artif. Intell.

View full text Add to dashboard Cite

Learning from unlabeled data, or self-learning, can substantially reduce the complexity of machine learning (ML) utilization in real-time deployment. While the development of un/semi-supervised algorithms shows promising results in learning with reduced labels, the fundamental assumption of data smoothness restricts its scope of application, especially with non-stationary data distributions in different domains. Leveraging trans-domain invariant causal relationships, causality has recently been employed to foster robust ML. In this paper, we have developed a generic method for self-labeling data that relies on known causality among interactive objects and learned temporal relations among causal events for identifying and associating labels and input data. The causal time-interval between asynchronous cause and effect events is studied to achieve self-labeling. We utilize dynamical system theory in a 1-d setting to demonstrate that our proposed method outperforms traditional feature similarity based semi-supervised learning. A computer simulation experiment was conducted to generate high-dimensional data, and the comprehensive results reveal the potential of learning adaptation in dynamic environments to improve ML robustness against shifts in data distribution.Impact Statement-Continual learning critically affects AI's adaptability in dynamic environments. Semi-supervised learning techniques which assume feature similarity and distribution smoothness are widely used to generate labels for unlabeled data samples to achieve adaptive learning, but are less effective in dynamic environments with non-stationary distributions. The causality-based self-labeling method introduced in this paper overcomes these limitations. A proof using dynamical systems theory has been applied to show the theoretical merits compared to traditional semi-supervised methods. A simulation experiment is conducted and significant accuracy improvement (over 10%) has been observed compared to several semi-supervised methods. This method offers an alternative way of continual and adaptive learning, leveraging causal knowledge in multimodal environments and applications.

show abstract

Section: Related Workmentioning

confidence: 99%

A Self-Labeling Method for Adaptive Machine Learning by Interactive Causality

Ren,

Yen,

2024

IEEE Trans. Artif. Intell.

View full text Add to dashboard Cite

show abstract

“…The time spans where causality is visible, might range from half a second to half a minute, occur several times, and can be interrupted by irrelevant scenes (e.g., one participant talking while the other participant is listening) that differ in the length of time. As outlined above, the direction of influence in a subinterval can either be bidirectional, or unidirectional driven by either S or R. This implies that three unwanted effects can occur, if the full time span is analysed: first, temporal relations are not found at all; second, bidirectional relations mask temporal unidirectional relations and; third, an unidirectional relation from X to Y masks temporal bidirectional influence or unidirectional influence from Y to X. Li et al (Li et al, 2017) give an example where temporal GC is not being detected, when the full time span is used for model fitting.…”

Section: Relevant Interval Selectionmentioning

confidence: 99%

Causal Inference in Nonverbal Dyadic Communication with Relevant Interval Selection and Granger Causality

Müller

Shadaydeh

Thümmel

et al. 2019

Proceedings of the 14th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Application

View full text Add to dashboard Cite

Human nonverbal emotional communication in dyadic dialogs is a process of mutual influence and adaptation. Identifying the direction of influence, or cause-effect relation between participants, is a challenging task due to two main obstacles. First, distinct emotions might not be clearly visible. Second, participants causeeffect relation is transient and variant over time. In this paper, we address these difficulties by using facial expressions that can be present even when strong distinct facial emotions are not visible. We also propose to apply a relevant interval selection approach prior to causal inference to identify those transient intervals where adaptation process occurs. To identify the direction of influence, we apply the concept of Granger causality to the time series of facial expressions on the set of relevant intervals. We tested our approach on synthetic data and then applied it to newly, experimentally obtained data. Here, we were able to show that a more sensitive facial expression detection algorithm and a relevant interval detection approach is most promising to reveal the cause-effect pattern for dyadic communication in various instructed interaction conditions. arXiv:1810.12171v1 [cs.CV] 29 Oct 2018 4. We propose a method for the selection of the relevant time intervals where GC should be applied, and show based on synthetic as well as real data, the superiority of the proposed method in detecting cause-effect relations when compared to applying GC on the full time series.

show abstract

“…Here, Granger causality does not empirically prove actual causation between events but acts as a stepping stone to explore the phenomenon relating two events participating in a cause-effect relationship. Granger-causal features have been discovered with rule-based analytics models [8] and feature-based analytics models [9]. Our approach to causal inference also builds a feature-based analytics model.…”

Section: Causal Inference In Time Series Analysismentioning

confidence: 99%

Discovering Granger-Causal Features from Deep Learning Networks

Chivukula

Liu

2018

Lecture Notes in Computer Science

View full text Add to dashboard Cite

In this research, we propose deep networks that discover Granger causes from multivariate temporal data generated in financial markets. We introduce a Deep Neural Network (DNN) and a Recurrent Neural Network (RNN) that discover Granger-causal features for bivariate regression on bivariate time series data distributions. These features are subsequently used to discover Grangercausal graphs for multivariate regression on multivariate time series data distributions. Our supervised feature learning process in proposed deep regression networks has favourable F-tests for feature selection and t-tests for model comparisons. The experiments, minimizing root mean squared errors in the regression analysis on real stock market data obtained from Yahoo Finance, demonstrate that our causal features significantly improve the existing deep learning regression models.

show abstract

Discovery of Causal Time Intervals

Cited by 12 publications

References 28 publications

A Self-Labeling Method for Adaptive Machine Learning by Interactive Causality

A Self-Labeling Method for Adaptive Machine Learning by Interactive Causality

Causal Inference in Nonverbal Dyadic Communication with Relevant Interval Selection and Granger Causality

Discovering Granger-Causal Features from Deep Learning Networks

Contact Info

Product

Resources

About