A Kernel Embedding-Based Approach for Nonstationary Causal Model Inference.

Although nonstationary data are more common in the real world, most existing causal discovery methods do not take nonstationarity into consideration. In this letter, we propose a kernel embedding-based approach, ENCI, for nonstationary causal model inference where data are collected from multiple domains with varying distributions. In ENCI, we transform the complicated relation of a cause-effect pair into a linear model of variables of which observations correspond to the kernel embeddings of the cause-and-effect distributions in different domains. In this way, we are able to estimate the causal direction by exploiting the causal asymmetry of the transformed linear model. Furthermore, we extend ENCI to causal graph discovery for multiple variables by transforming the relations among them into a linear nongaussian acyclic model. We show that by exploiting the nonstationarity of distributions, both cause-effect pairs and two kinds of causal graphs are identifiable under mild conditions. Experiments on synthetic and real-world data are conducted to justify the efficacy of ENCI over major existing methods.

Causal discovery via reproducing kernel Hilbert space embeddings.

Causal discovery via the asymmetry between the cause and the effect has proved to be a promising way to infer the causal direction from observations. The basic idea is to assume that the mechanism generating the cause distribution p(x) and that generating the conditional distribution p(y|x) correspond to two independent natural processes and thus p(x) and p(y|x) fulfill some sort of independence condition. However, in many situations, the independence condition does not hold for the anticausal direction; if we consider p(x, y) as generated via p(y)p(x|y), then there are usually some contrived mutual adjustments between p(y) and p(x|y). This kind of asymmetry can be exploited to identify the causal direction. Based on this postulate, in this letter, we define an uncorrelatedness criterion between p(x) and p(y|x) and, based on this uncorrelatedness, show asymmetry between the cause and the effect in terms that a certain complexity metric on p(x) and p(y|x) is less than the complexity metric on p(y) and p(x|y). We propose a Hilbert space embedding-based method EMD (an abbreviation for EMbeDding) to calculate the complexity metric and show that this method preserves the relative magnitude of the complexity metric. Based on the complexity metric, we propose an efficient kernel-based algorithm for causal discovery. The contribution of this letter is threefold. It allows a general transformation from the cause to the effect involving the noise effect and is applicable to both one-dimensional and high-dimensional data. Furthermore it can be used to infer the causal ordering for multiple variables. Extensive experiments on simulated and real-world data are conducted to show the effectiveness of the proposed method.

On Low-Rank Directed Acyclic Graphs and Causal Structure Learning.

Despite several advances in recent years, learning causal structures represented by directed acyclic graphs (DAGs) remains a challenging task in high-dimensional settings when the graphs to be learned are not sparse. In this article, we propose to exploit a low-rank assumption regarding the (weighted) adjacency matrix of a DAG causal model to help address this problem. We utilize existing low-rank techniques to adapt causal structure learning methods to take advantage of this assumption and establish several useful results relating interpretable graphical conditions to the low-rank assumption. Specifically, we show that the maximum rank is highly related to hubs, suggesting that scale-free (SF) networks, which are frequently encountered in practice, tend to be low rank. Our experiments demonstrate the utility of the low-rank adaptations for a variety of data models, especially with relatively large and dense graphs. Moreover, with a validation procedure, the adaptations maintain a superior or comparable performance even when graphs are not restricted to be low rank.

Causality in linear nongaussian acyclic models in the presence of latent gaussian confounders.

LiNGAM has been successfully applied to some real-world causal discovery problems. Nevertheless, causal sufficiency is assumed; that is, there is no latent confounder of the observations, which may be unrealistic for real-world problems. Taking into the consideration latent confounders will improve the reliability and accuracy of estimations of the real causal structures. In this letter, we investigate a model called linear nongaussian acyclic models in the presence of latent gaussian confounders (LiNGAM-GC) which can be seen as a specific case of lvLiNGAM. This model includes the latent confounders, which are assumed to be independent gaussian distributed and statistically independent of the disturbances. To tackle the causal discovery problem of this model, first we propose a pairwise cumulant-based measure of causal directions for cause-effect pairs. We prove that in spite of the presence of latent gaussian confounders, the causal direction of the observed cause-effect pair can be identified under the mild condition that the disturbances are simultaneously supergaussian or subgaussian. We propose a simple and efficient method to detect the violation of this condition. We extend our work to multivariate causal network discovery problems. Specifically we propose algorithms to estimate the causal network structure, including causal ordering and causal strengths, using an iterative root finding-removing scheme based on pairwise measure. To address the redundant edge problem due to the finite sample size effect, we develop an efficient bootstrapping-based pruning algorithm. Experiments on synthetic data and real-world data have been conducted to show the applicability of our model and the effectiveness of our proposed algorithms.

Face-specific negative bias of aesthetic perception in depression: Behavioral and EEG evidence.

Introduction: Symptoms of depression are associated with the dysfunction of neural systems such as the emotion, reward system, and the default mode network. These systems were suggested by the model of neuroaesthetics as key contributions to aesthetic experience, leading to the prediction of atypical aesthetic orientation in depression. Here we investigated the aesthetic characteristics of depression and the corresponding neural underpinnings. Methods: Fifty-two (25 depression patients, 27 healthy controls) participants were asked to make aesthetic judgments on faces and landscapes in an electroencephalographic (EEG) experiment. Results: Our results indicate that relative to the controls, the depression tended to give ugly judgments and refrained from giving beautiful judgments, which was observed only for faces but not for landscapes. We also found that the face-induced component N170 was more negative in the depression group than the control group for ugly and neutral faces. Moreover, the aesthetic evaluation of ugly faces was associated with decreased N200 negativity in the depression group than in the control group, while the evaluation of beautiful faces was associated with decreased brain synchronization at the theta band. Discussion: These results suggested a face-specific negative aesthetic bias in depression which can help to design and develop aesthetics-oriented schemes in assisting the clinical diagnosis and therapy of depression.

Prognostic value of the lymph node ratio in stage III colorectal cancer.

The nodal stage of colorectal cancer is based on the number of positive nodes. It is inevitably affected by the number of removed lymph nodes, but lymph node ratio can be unaffected. We investigated the value of lymph node ratio in stage III colorectal cancer in this study. The clinicopathologic factors and follow-up data of 145 cases of stage III colorectal cancer between January 1998 and December 2008 were analyzed retrospectively. The Pearson and Spearman correlation analyses were used to determine the correlation coefficient, the Kaplan-Meier method was used to analyze survival, and the Cox proportional hazard regression model was used for multivariate analysis in forward stepwise regression. We found that lymph node ratio was not correlated with the number of removed lymph nodes (r = -0.154, P = 0.065), but it was positively correlated with the number of positive lymph nodes (r = 0.739, P < 0.001) and N stage (r = 0.695, P < 0.001). Kaplan-Meier survival analysis revealed that tumor configuration, intestinal obstruction, serum carcinoembryonic antigen (CEA) concentration, T stage, N stage, and lymph node ratio were associated with disease-free survival of patients with stage III colorectal cancer (P < 0.05). Multivariate analysis showed that serum CEA concentration, T stage, and lymph node ratio were prognostic factors for disease-free survival (P < 0.05), whereas N stage failed to achieve significance (P = 0.664). We confirmed that lymph node ratio was a prognostic factor in stage III colorectal cancer and had a better prognostic value than did N stage.

Contrastive-ACE : Domain Generalization Through Alignment of Causal Mechanisms.

Domain generalization aims to learn knowledge invariant across different distributions while semantically meaningful for downstream tasks from multiple source domains, to improve the model's generalization ability on unseen target domains. The fundamental objective is to understand the underlying "invariance" behind these observational distributions and such invariance has been shown to have a close connection to causality. While many existing approaches make use of the property that causal features are invariant across domains, we consider the invariance of the average causal effect of the features to the labels. This invariance regularizes our training approach in which interventions are performed on features to enforce stability of the causal prediction by the classifier across domains. Our work thus sheds some light on the domain generalization problem by introducing invariance of the mechanisms into the learning process. Experiments on several benchmark datasets demonstrate the performance of the proposed method against SOTAs.

A natural marmoset model of genetic generalized epilepsy.

Epilepsy has been extensively studied as a common neurological disease. Efforts have been made on rodent and other animal models to reveal the pathogenic mechanisms of epilepsy and develop new drugs for treatment. However, the features of current epilepsy models cannot fully mimic different types of epilepsy in humans, hence non-human primate models of epilepsy are required. The common marmoset (Callithrix jacchus) is a New World monkey that is widely used to study brain function. Here, we present a natural marmoset model of generalized epilepsy. In this unique marmoset family, generalized epilepsy was successfully induced by handling operations in some individuals. We mapped the marmoset family with handling-sensitive epilepsy and found that the epileptic phenotype can be inherited. These marmosets were more sensitive to the epilepsy inducers pentylenetetrazol. Using electrocorticogram (ECoG) recordings, we detected epileptiform discharge in marmosets with a history of seizures. In summary, we report a family of marmosets with generalized seizures induced by handling operations. This epileptic marmoset family provides insights to better understand the mechanism of generalized epilepsy and helps to develop new therapeutic methods.

Decomposing Working Memory in Recurrent Major Depression: Impaired Encoding and Limited Maintenance Immune-to-Encoding Constraint.

It is generally believed that working memory (WM) is dysfunctional in depression. However, whether this impaired performance originates from impaired encoding, maintenance or both stages is still unclear. Here, we aimed to decompose the abnormal characteristics of encoding and maintenance in patients with recurrent major depressive disorder (MDD). Thirty patients and thirty-nine healthy controls completed a spatial working memory task where the encoding time and the retention time could vary under different load levels. Encoding performance was assessed by comparing accuracies between short and long encoding times, and maintenance performance was assessed by comparing accuracies between short and long retention times. The results show a lower performance in depression than the controls. However, while the decreased accuracy by long retention (vs. short retention) was increased by a short encoding time in the control group, the retention performance of the depression group did not further suffer from the short encoding time. The generally impaired encoding, together with limited maintenance of immunity against the constrained encoding time, suggests a common bias for fixed internal processing over external processing in recurrent MDD. The paradigm provided in this study can be a convenient and efficient clinical test for assessing the WM encoding and maintenance function.

Domain Generalization via Multidomain Discriminant Analysis.

Domain generalization (DG) aims to incorporate knowledge from multiple source domains into a single model that could generalize well on unseen target domains. This problem is ubiquitous in practice since the distributions of the target data may rarely be identical to those of the source data. In this paper, we propose Multidomain Discriminant Analysis (MDA) to address DG of classification tasks in general situations. MDA learns a domain-invariant feature transformation that aims to achieve appealing properties, including a minimal divergence among domains within each class, a maximal separability among classes, and overall maximal compactness of all classes. Furthermore, we provide the bounds on excess risk and generalization error by learning theory analysis. Comprehensive experiments on synthetic and real benchmark datasets demonstrate the effectiveness of MDA.