Search for: All records

As an alternative to resource-intensive deep learning approaches to the continual learning problem, we propose a simple, fast algorithm inspired by adaptive resonance theory (ART). To cope with the curse of dimensionality and avoid catastrophic forgetting, we apply incremental principal component analysis (IPCA) to the model's previously learned weights. Experiments show that this approach approximates the performance achieved using static PCA and is competitive with continual deep learning methods. Our implementation is available on https://github.com/neil-ash/ART-IPCA 

We consider the problem of learning predictive models from longitudinal data, consisting of irregularly repeated, sparse observations from a set of individuals over time. Such data of- ten exhibit longitudinal correlation (LC) (correlations among observations for each individual over time), cluster correlation (CC) (correlations among individuals that have similar char- acteristics), or both. These correlations are often accounted for using mixed effects models that include fixed effects and random effects, where the fixed effects capture the regression parameters that are shared by all individuals, whereas random effects capture those parameters that vary across individuals. However, the current state-of-the-art methods are unable to se- lect the most predictive fixed effects and random effects from a large number of variables, while accounting for complex cor- relation structure in the data and non-linear interactions among the variables. We propose Longitudinal Multi-Level Factoriza- tion Machine (LMLFM), to the best of our knowledge, the first model to address these challenges in learning predictive mod- els from longitudinal data. We establish the convergence prop- erties, and analyze the computational complexity, of LMLFM. We present results of experiments with both simulated and real-world longitudinal data which show that LMLFM out- performs the state-of-the-art methods in terms of predictive accuracy, variable selection ability, and scalability to data with large number of variables. The code and supplemental material is available at https://github.com/junjieliang672/LMLFM. 

Graph Neural Networks (GNN) offer the powerful approach to node classification in complex networks across many domains including social media, E-commerce, and FinTech. However, recent studies show that GNNs are vulnerable to attacks aimed at adversely impacting their node classification performance. Existing studies of adversarial attacks on GNN focus primarily on manipulating the connectivity between existing nodes, a task that requires greater effort on the part of the attacker in real-world applications. In contrast, it is much more expedient on the part of the attacker to inject adversarial nodes, e.g., fake profiles with forged links, into existing graphs so as to reduce the performance of the GNN in classifying existing nodes. Hence, we consider a novel form of node injection poisoning attacks on graph data. We model the key steps of a node injection attack, e.g., establishing links between the injected adversarial nodes and other nodes, choosing the label of an injected node, etc. by a Markov Decision Process. We propose a novel reinforcement learning method for Node Injection Poisoning Attacks (NIPA), to sequentially modify the labels and links of the injected nodes, without changing the connectivity between existing nodes. Specifically, we introduce a hierarchical Q-learning network to manipulate the labels of the adversarial nodes and their links with other nodes in the graph, and design an appropriate reward function to guide the reinforcement learning agent to reduce the node classification performance of GNN. The results of the experiments show that NIPA is consistently more effective than the baseline node injection attack methods for poisoning graph data on three benchmark datasets. 

Eliciting causal effects from interventions and observations is one of the central concerns of science, and increasingly, artificial intelligence. We provide an algorithm that, given a causal graph G, determines MIC(G), a minimum intervention cover of G, i.e., a minimum set of interventions that suffices for identifying every causal effect that is identifiable in a causal model characterized by G. We establish the completeness of do-calculus for computing MIC(G). MIC(G) effectively offers an efficient compilation of all of the information obtainable from all possible interventions in a causal model characterized by G. Minimum intervention cover finds applications in a variety of contexts including counterfactual inference, and generalizing causal effects across experimental settings. We analyze the computational complexity of minimum intervention cover and identify some special cases of practical interest in which MIC(G) can be computed in time that is polynomial in the size of G. 

We consider the problem of learning causal re- lationships from relational data. Existing ap- proaches rely on queries to a relational condi- tional independence (RCI) oracle to establish and orient causal relations in such a setting. In practice, queries to a RCI oracle have to be replaced by reliable tests for RCI against available data. Relational data present several unique challenges in testing for RCI. We study the conditions under which traditional iid-based CI tests yield reliable answers to RCI queries against relational data. We show how to conduct CI tests against relational data to robustly recover the underlying relational causal struc- ture. Results of our experiments demonstrate the effectiveness of our proposed approach. 

Data from many real-world applications can be nat- urally represented by multi-view networks where the different views encode different types of rela- tionships (e.g., friendship, shared interests in mu- sic, etc.) between real-world individuals or enti- ties. There is an urgent need for methods to ob- tain low-dimensional, information preserving and typically nonlinear embeddings of such multi-view networks. However, most of the work on multi- view learning focuses on data that lack a net- work structure, and most of the work on net- work embeddings has focused primarily on single- view networks. Against this background, we con- sider the multi-view network representation learn- ing problem, i.e., the problem of constructing low- dimensional information preserving embeddings of multi-view networks. Specifically, we investigate a novel Generative Adversarial Network (GAN) framework for Multi-View Network Embedding, namely MEGAN, aimed at preserving the informa- tion from the individual network views, while ac- counting for connectivity across (and hence com- plementarity of and correlations between) differ- ent views. The results of our experiments on two real-world multi-view data sets show that the em- beddings obtained using MEGAN outperform the state-of-the-art methods on node classification, link prediction and visualization tasks. 

Tests of conditional independence (CI) of ran- dom variables play an important role in ma- chine learning and causal inference. Of partic- ular interest are kernel-based CI tests which allow us to test for independence among ran- dom variables with complex distribution func- tions. The efficacy of a CI test is measured in terms of its power and its calibratedness. We show that the Kernel CI Permutation Test (KCIPT) suffers from a loss of calibratedness as its power is increased by increasing the number of bootstraps. To address this limita- tion, we propose a novel CI test, called Self- Discrepancy Conditional Independence Test (SDCIT). SDCIT uses a test statistic that is a modified unbiased estimate of maximum mean discrepancy (MMD), the largest difference in the means of features of the given sample and its permuted counterpart in the kernel-induced Hilbert space. We present results of experi- ments that demonstrate SDCIT is, relative to the other methods: (i) competitive in terms of its power and calibratedness, outperforming other methods when the number of condition- ing variables is large; (ii) more robust with re- spect to the choice of the kernel function; and (iii) competitive in run time.