Abstract--In Intelligent Tutoring System (ITS), tracing the student's knowledge state during learning has been studied for several decades in order to provide more supportive learning instructions. In this paper, we propose a novel model for knowledge tracing that i) captures students' learning ability and dynamically assigns students into distinct groups with similar ability at regular time intervals, and ii) combines this information with a Recurrent Neural Network architecture known as Deep Knowledge Tracing. Experimental results confirm that the proposed model is significantly better at predicting student performance than well known state-of-the-art techniques for student modelling. ITS is an active field of research that aims to provide personalized instructions to students. A wide array of Artificial Intelligence and Knowledge Representation techniques have been explored, of which we can mention rule-based and Bayesian representation of student knowledge and misconceptions, skills modeling with logistic regression in Item Response Theory, case-based reasoning, and, more recently reinforcement learning and deep learning [1], [2]. One can even argue that most of the main techniques found in Artificial Intelligence and Data Mining have found their way into the field of ITS, and in particular for the problem of knowledge tracing, which aims to model the student's state of mastery of conceptual or procedural knowledge from observed performance on tasks [3].

The inference of the causal relationship between a pair of observed variables is a fundamental problem in science, and most existing approaches are based on one single causal model. In practice, however, observations are often collected from multiple sources with heterogeneous causal models due to certain uncontrollable factors, which renders causal analysis results obtained by a single model skeptical. In this paper, we generalize the Additive Noise Model (ANM) to a mixture model, which consists of a finite number of ANMs, and provide the condition of its causal identifiability. To conduct model estimation, we propose Gaussian Process Partially Observable Model (GPPOM), and incorporate independence enforcement into it to learn latent parameter associated with each observation. Causal inference and clustering according to the underlying generating mechanisms of the mixture model are addressed in this work. Experiments on synthetic and real data demonstrate the effectiveness of our proposed approach.

Collaborative filtering (CF) has been successfully employed by many modern recommender systems. Conventional CF-based methods use the user-item interaction data as the sole information source to recommend items to users. However, CF-based methods are known for suffering from cold start problems and data sparsity problems. Hybrid models that utilize auxiliary information on top of interaction data have increasingly gained attention. A few "collaborative learning"-based models, which tightly bridges two heterogeneous learners through mutual regularization, are recently proposed for the hybrid recommendation. However, the "collaboration" in the existing methods are actually asynchronous due to the alternative optimization of the two learners. Leveraging the recent advances in variational autoencoder~(VAE), we here propose a model consisting of two streams of mutual linked VAEs, named variational collaborative model (VCM). Unlike the mutual regularization used in previous works where two learners are optimized asynchronously, VCM enables a synchronous collaborative learning mechanism. Besides, the two stream VAEs setup allows VCM to fully leverages the Bayesian probabilistic representations in collaborative learning. Extensive experiments on three real-life datasets have shown that VCM outperforms several state-of-art methods.

Developing automated methods to give advance warning of large gatherings of people, such as protests and social unrest events, are of interest to government agencies worldwide. With such events often being organised over online social media platforms, there exists the possibility to provide prior warning of large events solely through monitoring online data streams. Researchers have used open online data sources such as Twitter (Borge-Holthoefer et al., 2016; Agarwal and Sureka, 2016), Facebook, Tumblr (Xu et al., 2014), and Flickr (Alanyali et al., 2015) to characterise information propagation processes around protests, and have deployed machine learning methods on social media as well as blogs, news sources, and the dark web (Korkmaz et al., 2016) to predict civil unrest events. Twitter data in particular has been used broadly to monitor diverse largescale trends such as stock behaviour (Bollen et al., 2011), public opinion polling around 1 issues like climate change (Cody et al., 2015), and health characteristics (Alajajian et al., 2017). Recent studies have focussed on Twitter's role in particular in mobilisation and discourse around protest action in the United States (Theocharis et al., 2015; Gallagher et al., 2018).

Modern enterprise data science teams are technically diverse. The members of these teams differ in their education levels (bachelors, masters, PhD), majors (CS, statistics, natural sciences), prior work experiences (advertising, actuaries, finance, experimental physics) and used tools (Excel, SAS, R, Python, Java). One is more likely to find biology or computational neuroscience majors in data science teams today than in actuarial or financial firms of the past. But there is a more subtle difference than the ones listed above: team members with quantitative backgrounds may differ in exposure to the same set of concepts. Take the example of regression. Both engineering and statistics departments devote a portion of their curriculum to teaching line fitting.

Exemplar-based synthesis of geology using kernel discrepancies and generative neural networks Shing Chan and Ahmed H. Elsheikh Heriot-Watt University, United Kingdom School of Energy, Geoscience, Infrastructure and Society September 24, 2018 Abstract We propose a framework for synthesis of geological images based on an exemplar image (a.k.a. We synthesize new realizations such that the discrepancy in the patch distribution between the realizations and the exemplar image is minimized. Such discrepancy is quantified using a kernel method for two-sample test called maximum mean discrepancy. To enable fast synthesis, we train a generative neural network in an offline phase to sample realizations efficiently during deployment, while also providing a parametrization of the synthesis process. We assess the framework on a classical binary image representing channelized subsurface reservoirs, finding that the method reproduces the visual patterns and spatial statistics (image histogram and two-point probability functions) of the exemplar image. 1 Introduction A challenge in subsurface flow simulations is to obtain a complete and accurate image of subsurface properties, such as permeability and porosity, that are crucial for accurate flow predictions. Since it is virtually impossible to obtain direct measurements at every point of the domain under study, engineers can only rely on indirect estimations of the subsurface properties, e.g. from seismic images and sparse measurements obtained from wells. Traditionally, the properties are modeled based on their two-point statistics; however, this tends to produce images of the subsurface that are far from realistic. In many scenarios, such as in channelized systems where the properties follow an almost binary distribution and contain strong spatial correlations, two-point statistics are not enough to describe the distribution of the properties. This shortcoming led to the development of alternative algorithmic approaches to synthesize subsuface images that can capture multipoint statistics. These methods start from an exemplar image (also called training image in the geology literature) that is deemed representative of the subsurface under study, meaning that the spatial statistics in this image is believed to be similar to that of the subsurface.

We consider the clustering problem of attributed graphs. Our challenge is how we can design an effective and efficient clustering method that precisely captures the hidden relationship between the topology and the attributes in real-world graphs. We propose Non-linear Attributed Graph Clustering by Symmetric Non-negative Matrix Factorization with Positive Unlabeled Learning. The features of our method are three holds. 1) it learns a non-linear projection function between the different cluster assignments of the topology and the attributes of graphs so as to capture the complicated relationship between the topology and the attributes in real-world graphs, 2) it leverages the positive unlabeled learning to take the effect of partially observed positive edges into the cluster assignment, and 3) it achieves efficient computational complexity, $O((n^2+mn)kt)$, where $n$ is the vertex size, $m$ is the attribute size, $k$ is the number of clusters, and $t$ is the number of iterations for learning the cluster assignment. We conducted experiments extensively for various clustering methods with various real datasets to validate that our method outperforms the former clustering methods regarding the clustering quality.

Single and coupled matrix-tensor factorization (CMTF) has been widely used in this context for imputation-based recommendation from ratings, social network, and other user-item data. When this side information is in the form of item-item correlation matrices or graphs, existing CMTF algorithms may fall short. Alleviating current limitations, we introduce a novel model coined coupled graph-tensor factorization (CGTF) that judiciously accounts for graph-related side information. The CGTF model has the potential to overcome practical challenges, such as missing slabs from the tensor and/or missing rows/columns from the correlation matrices. A novel alternating direction method of multipliers (ADMM) is also developed that recovers the nonnegative factors of CGTF. Our algorithm enjoys closed-form updates that result in reduced computational complexity and allow for convergence claims. A novel direction is further explored by employing the interpretable factors to detect graph communities having the tensor as side information. The resulting community detection approach is successful even when some links in the graphs are missing. Results with real data sets corroborate the merits of the proposed methods relative to state-of-the-art competing factorization techniques in providing recommendations and detecting communities.

The goal is to distinguish whether the underlying graph is empty, i.e., the model consists of independent Rademacher variables, versus the alternative that the underlying graph contains a subgraph of a certain structure. We give matching upper and lower minimax bounds under which testing this problem is possible/impossible respectively. Our results reveal that a key quantity called graph arboricity drives the testability of the problem. On the computational front, under a conjecture of the computational hardness of sparse principal component analysis, we prove that, unless the signal is strong enough, there are no polynomial time linear tests on the sample covariance matrix which are capable of testing this problem.

Simulation plays an essential role in comprehending a target system in many fields of social and industrial sciences. A major task in simulation is the estimation of parameters, and optimal parameters to express the observed data need to directly elucidate the properties of the target system as the design of the simulator is based on the expert's domain knowledge. However, skilled human experts struggle to find the desired parameters.Data assimilation therefore becomes an unavoidable task in simulator design to reduce the cost of simulator optimization. Another necessary task is extrapolation; in many practical cases, the prediction based on simulation results will be often outside of the dominant range of the given data area, and this is referred to as the covariate shift. This paper focuses on the regression problem with the covariate shift. While the parameter estimation for the covariate shift has been studied thoroughly in parametric and nonparametric settings, conventional statistical methods of parameter searching are not applicable in the data assimilation of the simulation owing to the properties of the likelihood function: intractable or nondifferentiable. To address these problems, we propose a novel framework of Bayesian inference based on kernel mean embedding that comprises an extended kernel approximate Bayesian computation (ABC) of the importance weighted regression, kernel herding, and the kernel sum rule. This framework makes the prediction available in covariate shift situations, and its effectiveness is evaluated in both synthetic numerical experiments and a widely used production simulator.