Tsai-Ching Lu
Tsai-Ching Lu has a total of 11 co-authors. Here are the top co-authors by number of shared publications:
- Aruna Jammalamadaka
HRL Laboratories, LLC
2 shared publications - Sasha Strelnikoff
HRL Laboratories, LLC
2 shared publications Ana Morgenstern
University of Miami, Physics Department, USA
1 shared publication- Hong Qi
Peking University Third Hospital
1 shared publication Joseph Comer
HRL Laboratories, LLC
1 shared publication- Lei Zhang
HRL Laboratories, LLC
1 shared publication
Latest Publications
- Open Access
- 8 May 2023
- Published by University of Florida George A Smathers Libraries in The International FLAIRS Conference Proceedings
- Vol. 36
Semi-supervised Learning of Visual Causal Macrovariables
Discovery of causally related concepts is one of the key challenges in extracting knowledge from observational data. Lower-dimensional “causal macrovariables” represent concepts which preserve all relevant causal information in high-dimensional systems. Existing causal macrovariable discovery algorithms are limited by assumptions about known and controllable interventions. We propose a variational autoencoder-inspired architecture with regularization terms for semi-supervised causal macrovariable discovery. These terms impose domain knowledge regarding visual causal concepts to differentiate between correlation and causation. Experiments on both synthetic and real-world datasets with known causal dynamics show that our method can discover more concise and precise causal macrovariables than unsupervised methods.
- 8 May 2023
- Published by University of Florida George A Smathers Libraries in The International FLAIRS Conference Proceedings
- Vol. 36
Open AccessSemi-supervised Learning of Visual Causal Macrovariables
Discovery of causally related concepts is one of the key challenges in extracting knowledge from observational data. Lower-dimensional “causal macrovariables” represent concepts which preserve all relevant causal information in high-dimensional systems. Existing causal macrovariable discovery algorithms are limited by assumptions about known and controllable interventions. We propose a variational autoencoder-inspired architecture with regularization terms for semi-supervised causal macrovariable discovery. These terms impose domain knowledge regarding visual causal concepts to differentiate between correlation and causation. Experiments on both synthetic and real-world datasets with known causal dynamics show that our method can discover more concise and precise causal macrovariables than unsupervised methods.
- Open Access
- 8 May 2023
- Published by University of Florida George A Smathers Libraries in The International FLAIRS Conference Proceedings
- Vol. 36
Causanom: Anomaly Detection With Flexible Causal Graphs
Causality-based anomaly detection methods provide at least two significant theoretical benefits over purely statistical methods: 1. Improved robustness to non-anomalous out-of-distribution data, which implies a reduction in false-alarms; 2. A potential for failure localization due to the topological ordering of the causal graph. Recent studies have considered the utilization of causality-based methods for time series anomaly detection, however, these methods require the causal graph to be fixed; resultingly, such methods are not robust to incorrectly estimated causal graphs and are not able to natively model counterfactual scenarios. To address these limitations, we introduce Causanom: a graph-based encoder-decoder neural network for time series anomaly detection. Causanom utilizes a node conditional data-stream representation in conjunction with a weighted graph aggregation function in order to efficiently capture heterogeneous node dynamics whilst allowing for a flexible graphical structure. We show that Causanom can be trained along with auxiliary constraints in order to tune the causal graph and improve performance. Additionally, we show that Causanom can be used to produce counterfactual data, which we leverage to identify violated causal relationships. Using real and synthetic time series data respectively, we show that Causanom performs at least as well as state-of-the-art baselines in the anomaly detection task and outperforms existing methods in a causal attribution task.
- 8 May 2023
- Published by University of Florida George A Smathers Libraries in The International FLAIRS Conference Proceedings
- Vol. 36
Open AccessCausanom: Anomaly Detection With Flexible Causal Graphs
Causality-based anomaly detection methods provide at least two significant theoretical benefits over purely statistical methods: 1. Improved robustness to non-anomalous out-of-distribution data, which implies a reduction in false-alarms; 2. A potential for failure localization due to the topological ordering of the causal graph. Recent studies have considered the utilization of causality-based methods for time series anomaly detection, however, these methods require the causal graph to be fixed; resultingly, such methods are not robust to incorrectly estimated causal graphs and are not able to natively model counterfactual scenarios. To address these limitations, we introduce Causanom: a graph-based encoder-decoder neural network for time series anomaly detection. Causanom utilizes a node conditional data-stream representation in conjunction with a weighted graph aggregation function in order to efficiently capture heterogeneous node dynamics whilst allowing for a flexible graphical structure. We show that Causanom can be trained along with auxiliary constraints in order to tune the causal graph and improve performance. Additionally, we show that Causanom can be used to produce counterfactual data, which we leverage to identify violated causal relationships. Using real and synthetic time series data respectively, we show that Causanom performs at least as well as state-of-the-art baselines in the anomaly detection task and outperforms existing methods in a causal attribution task.
- 1 June 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 169-172
Context matters: Improving the uses of big data for forecasting civil unrest: Emerging phenomena and big data
Open Source Indicators (OSI) such as Google Trends (GT) promise to uncover the social dynamics associated with behavior that precede episodes of civil unrest. There are myriad reasons why societies may become unstable: Our analysis does not require or inquire the underlying reasons for discontent but instead takes into account differences associated with variegated social contexts. This paper examines instances of this volatile behavior and suggests a simple model for predicting civil unrest events using GT as an open source indicator (OSI). It grounds the possibilities for prediction on the fact that social processes occur within a particular social context. As such, paying attention to the particular signals associated from each country is an important moderator for any model keen on predicting cases of extreme social behavior such as civil unrest.
- 1 June 2013
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 169-172
Context matters: Improving the uses of big data for forecasting civil unrest: Emerging phenomena and big data
Open Source Indicators (OSI) such as Google Trends (GT) promise to uncover the social dynamics associated with behavior that precede episodes of civil unrest. There are myriad reasons why societies may become unstable: Our analysis does not require or inquire the underlying reasons for discontent but instead takes into account differences associated with variegated social contexts. This paper examines instances of this volatile behavior and suggests a simple model for predicting civil unrest events using GT as an open source indicator (OSI). It grounds the possibilities for prediction on the fact that social processes occur within a particular social context. As such, paying attention to the particular signals associated from each country is an important moderator for any model keen on predicting cases of extreme social behavior such as civil unrest.