Interactive Machine Learning (IML) is an iterative learning process that tightly couples a human with a machine learner, which is widely used by researchers and practitioners to effectively solve a wide variety of real-world application problems. Although recent years have witnessed the proliferation of IML in the field of visual analytics, most recent surveys either focus on a specific area of IML or aim to summarize a visualization field that is too generic for IML. In this paper, we systematically review the recent literature on IML and classify them into a task-oriented taxonomy built by us. We conclude the survey with a discussion of open challenges and research opportunities that we believe are inspiring for future work in IML.

Abstract--While neural networks for learning representation of multi-view data have been previously proposed as one of the state-of-the-art multi-view dimension reduction techniques, how to make the representation discriminative with only a small amount of labeled data is not well-studied. We introduce a semisupervised neural network model, named Multi-view Discriminative Neural Network (MDNN), for multi-view problems. MDNN finds nonlinear view-specific mappings by projecting samples to a common feature space using multiple coupled deep networks. It is capable of leveraging both labeled and unlabeled data to project multi-view data so that samples from different classes are separated and those from the same class are clustered together. It also uses the interview correlation between views to exploit the available information in both the labeled and unlabeled data. Extensive experiments conducted on four datasets demonstrate the effectiveness of the proposed algorithm for multi-view semisupervised learning. In many real-world problems, more than one set of features, referred to as views of the data, are available. For example, a web page can be represented by text data, images, and metadata. Multiple views can help improve the performance of many learning tasks because each view can provide information complementary to others, and learning using all views can maximally exploit the information available.

This paper presents a novel approach for detection of liver abnormalities in an automated manner using ultrasound images. For this purpose, we have implemented a machine learning model that can not only generate labels (normal and abnormal) for a given ultrasound image but it can also detect when its prediction is likely to be incorrect. The proposed model abstains from generating the label of a test example if it is not confident about its prediction. Such behavior is commonly practiced by medical doctors who, when given insufficient information or a difficult case, can chose to carry out further clinical or diagnostic tests before generating a diagnosis. However, existing machine learning models are designed in a way to always generate a label for a given example even when the confidence of their prediction is low. We have proposed a novel stochastic gradient based solver for the learning with abstention paradigm and use it to make a practical, state of the art method for liver disease classification. The proposed method has been benchmarked on a data set of approximately 100 patients from MINAR, Multan, Pakistan and our results show that the proposed scheme offers state of the art classification performance.

I describe an optimal control view of adversarial machine learning, where the dynamical system is the machine learner, the input are adversarial actions, and the control costs are defined by the adversary's goals to do harm and be hard to detect. This view encompasses many types of adversarial machine learning, including test-item attacks, training-data poisoning, and adversarial reward shaping. The view encourages adversarial machine learning researcher to utilize advances in control theory and reinforcement learning.

Abstract--Matrix factorization (MF) has been widely used to discover the low-rank structure and to predict the missing entries of data matrix. In many real-world learning systems, the data matrix can be very high-dimensional but sparse. This poses an imbalanced learning problem, since the scale of missing entries is usually much larger than that of observed entries, but they cannot be ignored due to the valuable negative signal. For efficiency concern, existing work typically applies a uniform weight on missing entries to allow a fast learning algorithm. However, this simplification will decrease modeling fidelity, resulting in suboptimal performance for downstream applications. In this work, we weight the missing data non-uniformly, and more generically, we allow any weighting strategy on the missing data. To address the efficiency challenge, we propose a fast learning method, for which the time complexity is determined by the number of observed entries in the data matrix, rather than the matrix size. The key idea is twofold: 1) we apply truncated SVD on the weight matrix to get a more compact representation of the weights, and 2) we learn MF parameters with element-wise alternating least squares (eALS) and memorize the key intermediate variables to avoid repeating computations that are unnecessary. We conduct extensive experiments on two recommendation benchmarks, demonstrating the correctness, efficiency, and effectiveness of our fast eALS method. Atrices are a common data structure to represent the relation between two types of entities in learning systems [1]-[3]. In relational learning, matrix factorization (MF) is a popular approach for dimension reduction by representing the rows (entities of one type) and columns (entities of another type) as two low-rank matrices. The optimization of dimension reduction is usually achieved by minimizing the reconstruction error between the low-rank model and the original data. Xiangnan He and Tat-Seng Chua are with the School of Computing, National University of Singapore, Singapore, 117417. Jinhui Tang is with the School of Computer Science and Engineering, Nanjing University of Science and Technology, Nanjing, Jiangsu, China, 210094.

Learning representation on graph plays a crucial role in numerous tasks of pattern recognition. Different from grid-shaped images/videos, on which local convolution kernels can be lattices, however, graphs are fully coordinate-free on vertices and edges. In this work, we propose a Gaussian-induced convolution (GIC) framework to conduct local convolution filtering on irregular graphs. Specifically, an edge-induced Gaussian mixture model is designed to encode variations of subgraph region by integrating edge information into weighted Gaussian models, each of which implicitly characterizes one component of subgraph variations. In order to coarsen a graph, we derive a vertex-induced Gaussian mixture model to cluster vertices dynamically according to the connection of edges, which is approximately equivalent to the weighted graph cut. We conduct our multi-layer graph convolution network on several public datasets of graph classification. The extensive experiments demonstrate that our GIC is effective and can achieve the state-of-the-art results.

This work explores adaptations of successful multi-armed bandits policies to the online contextual bandits scenario with binary rewards using binary classification algorithms such as logistic regression as black-box oracles. Some of these adaptations are achieved through bootstrapping or approximate bootstrapping, while others rely on other forms of randomness, resulting in more scalable approaches than previous works, and the ability to work with any type of classification algorithm. In particular, the Adaptive-Greedy algorithm shows a lot of promise, in many cases achieving better performance than upper confidence bound and Thompson sampling strategies, at the expense of more hyperparameters to tune.

The vicinal risk minimization (VRM) principle, first proposed by \citet{vapnik1999nature}, is an empirical risk minimization (ERM) variant that replaces Dirac masses with vicinal functions. Although there is strong numerical evidence showing that VRM outperforms ERM if appropriate vicinal functions are chosen, a comprehensive theoretical understanding of VRM is still lacking. In this paper, we study the generalization bounds for VRM. Our results support Vapnik's original arguments and additionally provide deeper insights into VRM. First, we prove that the complexity of function classes convolving with vicinal functions can be controlled by that of the original function classes under the assumption that the function class is composed of Lipschitz-continuous functions. Then, the resulting generalization bounds for VRM suggest that the generalization performance of VRM is also effected by the choice of vicinity function and the quality of function classes. These findings can be used to examine whether the choice of vicinal function is appropriate for the VRM-based learning setting. Finally, we provide a theoretical explanation for existing VRM models, e.g., uniform distribution-based models, Gaussian distribution-based models, and mixup models.

We investigate 1) the rate at which refined properties of the empirical risk---in particular, gradients---converge to their population counterparts in standard non-convex learning tasks, and 2) the consequences of this convergence for optimization. Our analysis follows the tradition of norm-based capacity control. We propose vector-valued Rademacher complexities as a simple, composable, and user-friendly tool to derive dimension-free uniform convergence bounds for gradients in non-convex learning problems. As an application of our techniques, we give a new analysis of batch gradient descent methods for non-convex generalized linear models and non-convex robust regression, showing how to use any algorithm that finds approximate stationary points to obtain optimal sample complexity, even when dimension is high or possibly infinite and multiple passes over the dataset are allowed. Moving to non-smooth models we show----in contrast to the smooth case---that even for a single ReLU it is not possible to obtain dimension-independent convergence rates for gradients in the worst case. On the positive side, it is still possible to obtain dimension-independent rates under a new type of distributional assumption.

In this paper, we consider first-order algorithms for solving a class of non-convex non-concave min-max saddle-point problems, whose objective function is weakly convex (resp. weakly concave) in terms of the variable of minimization (resp. maximization). It has many important applications in machine learning, statistics, and operations research. One such example that attracts tremendous attention recently in machine learning is training Generative Adversarial Networks. We propose an algorithmic framework motivated by the inexact proximal point method, which solves the weakly monotone variational inequality corresponding to the original min-max problem by approximately solving a sequence of strongly monotone variational inequalities constructed by adding a strongly monotone mapping to the original gradient mapping. In this sequence, each strongly monotone variational inequality is defined with a proximal center that is updated using the approximate solution of the previous variational inequality. Our algorithm generates a sequence of solution that provably converges to a nearly stationary solution of the original min-max problem. The proposed framework is flexible because various subroutines can be employed for solving the strongly monotone variational inequalities. The overall computational complexities of our methods are established when the employed subroutines are subgradient method, stochastic subgradient method, gradient descent method and Nesterov's accelerated method and variance reduction methods for a Lipschitz continuous operator. To the best of our knowledge, this is the first work that establishes the non-asymptotic convergence to a nearly stationary point of a non-convex non-concave min-max problem.