A framework is introduced for actively and adaptively solving a sequence of machine learning problems, which are changing in bounded manner from one time step to the next. An algorithm is developed that actively queries the labels of the most informative samples from an unlabeled data pool, and that adapts to the change by utilizing the information acquired in the previous steps. Our analysis shows that the proposed active learning algorithm based on stochastic gradient descent achieves a near-optimal excess risk performance for maximum likelihood estimation. Furthermore, an estimator of the change in the learning problems using the active learning samples is constructed, which provides an adaptive sample size selection rule that guarantees the excess risk is bounded for sufficiently large number of time steps. Experiments with synthetic and real data are presented to validate our algorithm and theoretical results.

In this paper, we introduce a new form of amortized variational inference by using the forward KL divergence in a joint-contrastive variational loss. The resulting forward amortized variational inference is a likelihood-free method as its gradient can be sampled without bias and without requiring any evaluation of either the model joint distribution or its derivatives. We prove that our new variational loss is optimized by the exact posterior marginals in the fully factorized mean-field approximation, a property that is not shared with the more conventional reverse KL inference. Furthermore, we show that forward amortized inference can be easily marginalized over large families of latent variables in order to obtain a marginalized variational posterior. We consider two examples of variational marginalization. In our first example we train a Bayesian forecaster for predicting a simplified chaotic model of atmospheric convection. In the second example we train an amortized variational approximation of a Bayesian optimal classifier by marginalizing over the model space. The result is a powerful meta-classification network that can solve arbitrary classification problems without further training.

We study the effect of imperfect training data labels on the performance of classification methods. In a general setting, where the probability that an observation in the training dataset is mislabelled may depend on both the feature vector and the true label, we bound the excess risk of an arbitrary classifier trained with imperfect labels in terms of its excess risk for predicting a noisy label. This reveals conditions under which a classifier trained with imperfect labels remains consistent for classifying uncorrupted test data points. Furthermore, under stronger conditions, we derive detailed asymptotic properties for the popular $k$-nearest neighbour ($k$nn), Support Vector Machine (SVM) and Linear Discriminant Analysis (LDA) classifiers. One consequence of these results is that the $k$nn and SVM classifiers are robust to imperfect training labels, in the sense that the rate of convergence of the excess risks of these classifiers remains unchanged; in fact, it even turns out that in some cases, imperfect labels may improve the performance of these methods. On the other hand, the LDA classifier is shown to be typically inconsistent in the presence of label noise unless the prior probabilities of each class are equal. Our theoretical results are supported by a simulation study.

In this work, a novel sequential Monte Carlo filter is introduced which aims at efficient sampling of high-dimensional state spaces with a limited number of particles. Particles are pushed forward from the prior to the posterior density using a sequence of mappings that minimizes the Kullback-Leibler divergence between the posterior and the sequence of intermediate densities. The sequence of mappings represents a gradient flow. A key ingredient of the mappings is that they are embedded in a reproducing kernel Hilbert space, which allows for a practical and efficient algorithm. The embedding provides a direct means to calculate the gradient of the Kullback-Leibler divergence leading to quick convergence using well-known gradient-based stochastic optimization algorithms. Evaluation of the method is conducted in the chaotic Lorenz-63 system, the Lorenz-96 system, which is a coarse prototype of atmospheric dynamics, and an epidemic model that describes cholera dynamics. No resampling is required in the mapping particle filter even for long recursive sequences. The number of effective particles remains close to the total number of particles in all the experiments.

This paper considers the problem of removing costly features from a Bayesian network classifier. We want the classifier to be robust to these changes, and maintain its classification behavior. To this end, we propose a closeness metric between Bayesian classifiers, called the expected classification agreement (ECA). Our corresponding trimming algorithm finds an optimal subset of features and a new classification threshold that maximize the expected agreement, subject to a budgetary constraint. It utilizes new theoretical insights to perform branch-and-bound search in the space of feature sets, while computing bounds on the ECA. Our experiments investigate both the runtime cost of trimming and its effect on the robustness and accuracy of the final classifier.

Off-policy reinforcement learning enables near-optimal policy from suboptimal experience, thereby provisions opportunity for artificial intelligence applications in healthcare. Previous works have mainly framed patient-clinician interactions as Markov decision processes, while true physiological states are not necessarily fully observable from clinical data. We capture this situation with partially observable Markov decision process, in which an agent optimises its actions in a belief represented as a distribution of patient states inferred from individual history trajectories. A Gaussian mixture model is fitted for the observed data. Moreover, we take into account the fact that nuance in pharmaceutical dosage could presumably result in significantly different effect by modelling a continuous policy through a Gaussian approximator directly in the policy space, i.e. the actor. To address the challenge of infinite number of possible belief states which renders exact value iteration intractable, we evaluate and plan for only every encountered belief, through heuristic search tree by tightly maintaining lower and upper bounds of the true value of belief. We further resort to function approximations to update value bounds estimation, i.e. the critic, so that the tree search can be improved through more compact bounds at the fringe nodes that will be back-propagated to the root. Both actor and critic parameters are learned via gradient-based approaches. Our proposed policy trained from real intensive care unit data is capable of dictating dosing on vasopressors and intravenous fluids for sepsis patients that lead to the best patient outcomes.

We show that when a third party, the adversary, steps into the two-party setting (agent and operator) of safely interruptible reinforcement learning, a trade-off has to be made between the probability of following the optimal policy in the limit, and the probability of escaping a dangerous situation created by the adversary. So far, the work on safely interruptible agents has assumed a perfect perception of the agent about its environment (no adversary), and therefore implicitly set the second probability to zero, by explicitly seeking a value of one for the first probability. We show that (1) agents can be made both interruptible and adversary-resilient, and (2) the interruptibility can be made safe in the sense that the agent itself will not seek to avoid it. We also solve the problem that arises when the agent does not go completely greedy, i.e. issues with safe exploration in the limit. Resilience to perturbed perception, safe exploration in the limit, and safe interruptibility are the three pillars of what we call \emph{virtuously safe reinforcement learning}.

While knowledge representation and reasoning are considered the keys for human-level artificial intelligence, connectionist networks have been shown successful in a broad range of applications due to their capacity for robust learning and flexible inference under uncertainty. The idea of representing symbolic knowledge in connectionist networks has been well-received and attracted much attention from research community as this can establish a foundation for integration of scalable learning and sound reasoning. In previous work, there exist a number of approaches that map logical inference rules with feed-forward propagation of artificial neural networks (ANN). However, the discriminative structure of an ANN requires the separation of input/output variables which makes it difficult for general reasoning where any variables should be inferable. Other approaches address this issue by employing generative models such as symmetric connectionist networks, however, they are difficult and convoluted. In this paper we propose a novel method to represent propositional formulas in restricted Boltzmann machines which is less complex, especially in the cases of logical implications and Horn clauses. An integration system is then developed and evaluated in real datasets which shows promising results.

Inspired by how humans summarize long documents, we propose an accurate and fast summarization model that first selects salient sentences and then rewrites them abstractively (i.e., compresses and paraphrases) to generate a concise overall summary. We use a novel sentence-level policy gradient method to bridge the non-differentiable computation between these two neural networks in a hierarchical way, while maintaining language fluency. Empirically, we achieve the new state-of-the-art on all metrics (including human evaluation) on the CNN/Daily Mail dataset, as well as significantly higher abstractiveness scores. Moreover, by first operating at the sentence-level and then the word-level, we enable parallel decoding of our neural generative model that results in substantially faster (10-20x) inference speed as well as 4x faster training convergence than previous long-paragraph encoder-decoder models. We also demonstrate the generalization of our model on the test-only DUC-2002 dataset, where we achieve higher scores than a state-of-the-art model.

Technical analysis is used to discover investment opportunities. To test this hypothesis we propose an hybrid system using machine learning techniques together with genetic algorithms. Using technical analysis there are more ways to represent a currency exchange time series than the ones it is possible to test computationally, i.e., it is unfeasible to search the whole input feature space thus a genetic algorithm is an alternative. In this work, an architecture for automatic feature selection is proposed to optimize the cross validated performance estimation of a Naive Bayes model using a genetic algorithm. The proposed architecture improves the return on investment of the unoptimized system from 0,43% to 10,29% in the validation set. The features selected and the model decision boundary are visualized using the algorithm t-Distributed Stochastic Neighbor embedding.