Tu Dinh Nguyen

Learning vector representation of medical objects via EMR-driven nonnegative restricted Boltzmann machines (eNRBM)

Electronic medical record (EMR) offers promises for novel analytics. However, manual feature engineering from EMR is labor intensive because EMR is complex - it contains temporal, mixed-type and multimodal data packed in irregular episodes. We present a computational framework to harness EMR with minimal human supervision via restricted Boltzmann machine (RBM). The framework derives a new representation of medical objects by embedding them in a low-dimensional vector space. This new representation facilitates algebraic and statistical manipulations such as projection onto 2D plane (thereby offering intuitive visualization), object grouping (hence enabling automated phenotyping), and risk stratification. To enhance model interpretability, we introduced two constraints into model parameters: (a) nonnegative coefficients, and (b) structural smoothness. These result in a novel model called eNRBM (EMR-driven nonnegative RBM). We demonstrate the capability of the eNRBM on a cohort of 7578 mental health patients under suicide risk assessment. The derived representation not only shows clinically meaningful feature grouping but also facilitates short-term risk stratification. The F-scores, 0.21 for moderate-risk and 0.36 for high-risk, are significantly higher than those obtained by clinicians and competitive with the results obtained by support vector machines.

Latent Patient Profile Modelling and Applications with Mixed-Variate Restricted Boltzmann Machine

Efficient management of chronic diseases is critical in modern health care. We consider diabetes mellitus, and our ongoing goal is to examine how machine learning can deliver information for clinical efficiency. The challenge is to aggregate highly heterogeneous sources including demographics, diagnoses, pathologies and treatments, and extract similar groups so that care plans can be designed. To this end, we extend our recent model, the mixed-variate restricted Boltzmann machine (MV.RBM), as it seamlessly integrates multiple data types for each patient aggregated over time and outputs a homogeneous representation called “latent profile” that can be used for patient clustering, visualisation, disease correlation analysis and prediction. We demonstrate that the method outperforms all baselines on these tasks - the primary characteristics of patients in the same groups are able to be identified and the good result can be achieved for the diagnosis codes prediction.

Large-scale online kernel learning with random feature reparameterization

A typical online kernel learning method faces two fundamental issues: the complexity in dealing with a huge number of observed data points (a.k.a the curse of kernelization) and the difficulty in learning kernel parameters, which often assumed to be fixed. Random Fourier feature is a recent and effective approach to address the former by approximating the shift-invariant kernel function via Bocher’s theorem, and allows the model to be maintained directly in the random feature space with a fixed dimension, hence the model size remains constant w.r.t. data size. We further introduce in this paper the reparameterized random feature (RRF), a random feature framework for large-scale online kernel learning to address both aforementioned challenges. Our initial intuition comes from the so-called ‘reparameterization trick’ [Kingma and Welling, 2014] to lift the source of randomness of Fourier components to another space which can be independently sampled, so that stochastic gradient of the kernel parameters can be analytically derived. We develop a well-founded underlying theory for our method, including a general way to reparameterize the kernel, and a new tighter error bound on the approximation quality. This view further inspires a direct application of stochastic gradient descent for updating our model under an online learning setting. We then conducted extensive experiments on several large-scale datasets where we demonstrate that our work achieves state-of-the-art performance in both learning efficacy and efficiency.

Learning sparse latent representation and distance metric for image retrieval

The performance of image retrieval depends critically on the semantic representation and the distance function used to estimate the similarity of two images. A good representation should integrate multiple visual and textual (e.g., tag) features and offer a step closer to the true semantics of interest (e.g., concepts). As the distance function operates on the representation, they are interdependent, and thus should be addressed at the same time. We propose a probabilistic solution to learn both the representation from multiple feature types and modalities and the distance metric from data. The learning is regularised so that the learned representation and information-theoretic metric will (i) preserve the regularities of the visual/textual spaces, (ii) enhance structured sparsity, (iii) encourage small intra-concept distances, and (iv) keep inter-concept images separated. We demonstrate the capacity of our method on the NUS-WIDE data. For the well-studied 13 animal subset, our method outperforms state-of-the-art rivals. On the subset of single-concept images, we gain 79:5% improvement over the standard nearest neighbours approach on the MAP score, and 45.7% on the NDCG.

Distributed data augmented support vector machine on Spark

Support vector machines (SVMs) are widely-used for classification in machine learning and data mining tasks. However, they traditionally have been applied to small to medium datasets. Recent need to scale up with data size has attracted research attention to develop new methods and implementation for SVM to perform tasks at scale. Distributed SVMs are relatively new and studied recently, but the distributed implementation for SVM with data augmentation has not been developed. This paper introduces a distributed data augmentation implementation for SVM on Apache Spark, a recent advanced and popular platform for distributed computing that has been employed widely in research as well as in industry. We term our implementation sparkling vector machine (SkVM) which supports both classification and regression tasks by scanning through the data exactly once. In addition, we further develop a framework to handle the data with new classes arriving under an online classification setting where new data points can have labels that have not previously seen - a problem we term label-drift classification. We demonstrate the scalability of our proposed method on large-scale datasets with more than one hundred million data points. The experimental results show that the predictive performances of our method are comparable or better than those of baselines whilst the execution time is much faster at an order of magnitude.

Graph-induced restricted Boltzmann machines for document modeling

Discovering knowledge from unstructured texts is a central theme in data mining and machine learning. We focus on fast discovery of thematic structures from a corpus. Our approach is based on a versatile probabilistic formulation - the restricted Boltzmann machine (RBM) - where the underlying graphical model is an undirected bipartite graph. Inference is efficient - document representation can be computed with a single matrix projection, making RBMs suitable for massive text corpora available today. Standard RBMs, however, operate on bag-of-words assumption, ignoring the inherent underlying relational structures among words. This results in less coherent word thematic grouping. We introduce graph-based regularization schemes that exploit the linguistic structures, which in turn can be constructed from either corpus statistics or domain knowledge. We demonstrate that the proposed technique improves the group coherence, facilitates visualization, provides means for estimation of intrinsic dimensionality, reduces overfitting, and possibly leads to better classification accuracy.

Stabilizing High-Dimensional Prediction Models Using Feature Graphs

We investigate feature stability in the context of clinical prognosis derived from high-dimensional electronic medical records. To reduce variance in the selected features that are predictive, we introduce Laplacian-based regularization into a regression model. The Laplacian is derived on a feature graph that captures both the temporal and hierarchic relations between hospital events, diseases, and interventions. Using a cohort of patients with heart failure, we demonstrate better feature stability and goodness-of-fit through feature graph stabilization.

Energy-Based Localized Anomaly Detection in Video Surveillance

Automated detection of abnormal events in video surveillance is an important task in research and practical applications. This is, however, a challenging problem due to the growing collection of data without the knowledge of what to be defined as “abnormal”, and the expensive feature engineering procedure. In this paper we introduce a unified framework for anomaly detection in video based on the restricted Boltzmann machine (\(\text {RBM}\)), a recent powerful method for unsupervised learning and representation learning. Our proposed system works directly on the image pixels rather than hand-crafted features, it learns new representations for data in a completely unsupervised manner without the need for labels, and then reconstructs the data to recognize the locations of abnormal events based on the reconstruction errors. More importantly, our approach can be deployed in both offline and streaming settings, in which trained parameters of the model are fixed in offline setting whilst are updated incrementally with video data arriving in a stream. Experiments on three publicly benchmark video datasets show that our proposed method can detect and localize the abnormalities at pixel level with better accuracy than those of baselines, and achieve competitive performance compared with state-of-the-art approaches. Moreover, as RBM belongs to a wider class of deep generative models, our framework lays the groundwork towards a more powerful deep unsupervised abnormality detection framework.

One-Pass Logistic Regression for Label-Drift and Large-Scale Classification on Distributed Systems

Logistic regression (LR) for classification is the workhorse in industry, where a set of predefined classes is required. The model, however, fails to work in the case where the class labels are not known in advance, a problem we term label-drift classification. Label-drift classification problem naturally occurs in many applications, especially in the context of streaming settings where the incoming data may contain samples categorized with new classes that have not been previously seen. Additionally, in the wave of big data, traditional LR methods may fail due to their expense of running time. In this paper, we introduce a novel variant of LR, namely one-pass logistic regression (OLR) to offer a principled treatment for label-drift and large-scale classifications. To handle largescale classification for big data, we further extend our OLR to a distributed setting for parallelization, termed sparkling OLR (Spark-OLR). We demonstrate the scalability of our proposed methods on large-scale datasets with more than one hundred million data points. The experimental results show that the predictive performances of our methods are comparable orbetter than those of state-of-the-art baselines whilst the executiontime is much faster at an order of magnitude. In addition, the OLR and Spark-OLR are invariant to data shuffling and have no hyperparameter to tune that significantly benefits data practitioners and overcomes the curse of big data cross-validationto select optimal hyperparameters.

GoGP: Fast Online Regression with Gaussian Processes

One of the most current challenging problems in Gaussian process regression (GPR) is to handle large-scale datasets and to accommodate an online learning setting where data arrive irregularly on the fly. In this paper, we introduce a novel online Gaussian process model that could scale with massive datasets. Our approach is formulated based on alternative representation of the Gaussian process under geometric and optimization views, hence termed geometric-based online GP (GoGP). We developed theory to guarantee that with a good convergence rate our proposed algorithm always produces a (sparse) solution which is close to the true optima to any arbitrary level of approximation accuracy specified a priori. Furthermore, our method is proven to scale seamlessly not only with large-scale datasets, but also to adapt accurately with streaming data. We extensively evaluated our proposed model against state-of-the-art baselines using several large-scale datasets for online regression task. The experimental results show that our GoGP delivered comparable, or slightly better, predictive performance while achieving a magnitude of computational speedup compared with its rivals under online setting. More importantly, its convergence behavior is guaranteed through our theoretical analysis, which is rapid and stable while achieving lower errors.