Survival analysis is a hotspot in statistical research for modeling time-to-event information with data censorship handling, which has been widely used in many applications such as clinical research, information system and other fields with survivorship bias. Many works have been proposed for survival analysis ranging from traditional statistic methods to machine learning models. However, the existing methodologies either utilize counting-based statistics on the segmented data, or have a pre-assumption on the event probability distribution w.r.t. time. Moreover, few works consider sequential patterns within the feature space. In this paper, we propose a Deep Recurrent Survival Analysis model which combines deep learning for conditional probability prediction at fine-grained level of the data, and survival analysis for tackling the censorship. By capturing the time dependency through modeling the conditional probability of the event for each sample, our method predicts the likelihood of the true event occurrence and estimates the survival rate over time, i.e., the probability of the non-occurrence of the event, for the censored data. Meanwhile, without assuming any specific form of the event probability distribution, our model shows great advantages over the previous works on fitting various sophisticated data distributions. In the experiments on the three real-world tasks from different fields, our model significantly outperforms the state-of-the-art solutions under various metrics.

With an ever growing number of extractive summarization techniques being proposed, there is less clarity then ever about how good each system is compared to the rest. Several studies highlight the variance in performance of these systems with change in datasets or even across documents within the same corpus. An effective way to counter this variance and to make the systems more robust could be to use inputs from multiple systems when generating a summary. In the present work, we define a novel way of creating such ensemble by exploiting similarity between the content of candidate summaries to estimate their reliability. We define GlobalRank which captures the performance of a candidate system on an overall corpus and LocalRank which estimates its performance on a given document cluster. We then use these two scores to assign a weight to each individual systems, which is then used to generate the new aggregate ranking. Experiments on DUC2003 and DUC 2004 datasets show a significant improvement in terms of ROUGE score, over existing sate-of-art techniques.

The grey rectangle starts on the left of the screen and can be moved with two actions, "Right" and "Jump". The goal of this game is to reach the right of the screen while avoiding the white obstacle. There is only one specific distance (measured in number of pixels) to the obstacle where the agent has to chose the action "Jump" in order to pass over the obstacle. If jumping is chosen at any other point, the agent will inevitably crash into the obstacle. A reward of 1 is granted anytime the agent moves one pixel to the right (even in the air). The episode terminates if the agent reaches the right of the screen or touches the obstacle. We build a set of related tasks by varying two factors: the floor height and the position of the obstacle on the floor. The resulting set contains 1271 tasks. We use 6 of those for training and evaluate the generalization performance as the fraction of the remaining 1265 tasks the agent can solve.

We propose an unsupervised method to obtain cross-lingual embeddings without any parallel data or pre-trained word embeddings. The proposed model, which we call multilingual neural language models, takes sentences of multiple languages as an input. The proposed model contains bidirectional LSTMs that perform as forward and backward language models, and these networks are shared among all the languages. The other parameters, i.e. word embeddings and linear transformation between hidden states and outputs, are specific to each language. The shared LSTMs can capture the common sentence structure among all languages. Accordingly, word embeddings of each language are mapped into a common latent space, making it possible to measure the similarity of words across multiple languages. We evaluate the quality of the cross-lingual word embeddings on a word alignment task. Our experiments demonstrate that our model can obtain cross-lingual embeddings of much higher quality than existing unsupervised models when only a small amount of monolingual data (i.e.

MixUp (Zhang et al. 2017), a data augmentation approach through mixing random samples, has been shown to be able to significantly improve the predictive accuracy of the current art of deep neural networks. The power of MixUp, however, is mostly established empirically and its working and effectiveness have not been explained in any depth. In this paper, we develop a theoretical understanding for MixUp as a form of out-of-manifold regularization, which constrains the model on the input space beyond the data manifold. This analytical study also enables us to identify MixUp's limitation caused by manifold intrusion, where synthetic samples collide with real examples of the manifold. Such intrusion gives rise to over regularization and thereby under-fitting. To address this issue, we further propose a novel regularizer, where mixing policies are adaptively learned from the data and a manifold intrusion loss is embraced as to avoid collision with the data manifold. We empirically show, using several benchmark datasets, our regularizer's effectiveness in terms of over regularization avoiding and accuracy improvement upon current art of deep classification models and MixUp.

Deep reinforcement learning has obtained significant breakthroughs in recent years. Most methods in deep-RL achieve good results via the maximization of the reward signal provided by the environment, typically in the form of discounted cumulative returns. Such reward signals represent the immediate feedback of a particular action performed by an agent. However, tasks with sparse reward signals are still challenging to on-policy methods. In this paper, we introduce an effective characterization of past reward statistics (which can be seen as long-term feedback signals) to supplement this immediate reward feedback. In particular, value functions are learned with multi-critics supervision, enabling complex value functions to be more easily approximated in on-policy learning, even when the reward signals are sparse. We also introduce a novel exploration mechanism called "hot-wiring" that can give a boost to seemingly trapped agents. We demonstrate the effectiveness of our advantage actor multi-critic (A2MC) method across the discrete domains in Atari games as well as continuous domains in the MuJoCo environments. A video demo is provided at https://youtu.be/zBmpf3Yz8tc.

Monte Carlo Tree Search (MCTS) is particularly adapted to domains where the potential actions can be represented as a tree of sequential decisions. For an effective action selection, MCTS performs many simulations to build a reliable tree representation of the decision space. As such, a bottleneck to MCTS appears when enough simulations cannot be performed between action selections. This is particularly highlighted in continuously running tasks, for which the time available to perform simulations between actions tends to be limited due to the environment's state constantly changing. In this paper, we present an approach that takes advantage of the anytime characteristic of MCTS to increase the simulation time when allowed. Our approach is to effectively balance the prospect of selecting an action with the time that can be spared to perform MCTS simulations before the next action selection. For that, we considered the simulation time as a decision variable to be selected alongside an action. We extended the Hierarchical Optimistic Optimization applied to Tree (HOOT) method to adapt our approach to environments with a continuous decision space. We evaluated our approach for environments with a continuous decision space through OpenAI gym's Pendulum and Continuous Mountain Car environments and for environments with discrete action space through the arcade learning environment (ALE) platform. The evaluation results show that, with variable simulation times, the proposed approach outperforms the conventional MCTS in the evaluated continuous decision space tasks and improves the performance of MCTS in most of the ALE tasks.

Populating ontology graphs represents a long-standing problem for the Semantic Web community. Recent advances in translation-based graph embedding methods for populating instance-level knowledge graphs lead to promising new approaching for the ontology population problem. However, unlike instance-level graphs, the majority of relation facts in ontology graphs come with comprehensive semantic relations, which often include the properties of transitivity and symmetry, as well as hierarchical relations. These comprehensive relations are often too complex for existing graph embedding methods, and direct application of such methods is not feasible. Hence, we propose On2Vec, a novel translation-based graph embedding method for ontology population. On2Vec integrates two model components that effectively characterize comprehensive relation facts in ontology graphs. The first is the Component-specific Model that encodes concepts and relations into low-dimensional embedding spaces without a loss of relational properties; the second is the Hierarchy Model that performs focused learning of hierarchical relation facts. Experiments on several well-known ontology graphs demonstrate the promising capabilities of On2Vec in predicting and verifying new relation facts. These promising results also make possible significant improvements in related methods.

Training a neural network for a classification task typically assumes that the data to train are given from the beginning. However, in the real world, additional data accumulate gradually and the model requires additional training without accessing the old training data. This usually leads to the catastrophic forgetting problem which is inevitable for the traditional training methodology of neural networks. In this paper, we propose a memory-based continual learning method that is able to learn additional tasks while retaining the performance of previously learned tasks. Composed of two complementary networks, the Hippocampus-Net (H-Net) and the Cortex-Net (C-Net), our model estimates the index of the corresponding task for an input sample and utilizes a particular portion of itself with the estimated index. The C-Net guarantees no degradation in the performance of the previously learned tasks and the H-Net shows high confidence in finding the origin of an input sample.

Graph representation learning aims at transforming graph data into meaningful low-dimensional vectors to facilitate the employment of machine learning and data mining algorithms designed for general data. Most current graph representation learning approaches are transductive, which means that they require all the nodes in the graph are known when learning graph representations and these approaches cannot naturally generalize to unseen nodes. In this paper, we present a Fast Inductive Graph Representation Learning framework (FI-GRL) to learn nodes' low-dimensional representations. Our approach can obtain accurate representations for seen nodes with provable theoretical guarantees and can easily generalize to unseen nodes. Specifically, in order to explicitly decouple nodes' relations expressed by the graph, we transform nodes into a randomized subspace spanned by a random projection matrix. This stage is guaranteed to preserve the projection-cost of the normalized random walk matrix which is highly related to the normalized cut of the graph. Then feature extraction is achieved by conducting singular value decomposition on the obtained matrix sketch. By leveraging the property of projection-cost preservation on the matrix sketch, the obtained representation result is nearly optimal. To deal with unseen nodes, we utilize folding-in technique to learn their meaningful representations. Empirically, when the amount of seen nodes are larger than that of unseen nodes, FI-GRL always achieves excellent results. Our algorithm is fast, simple to implement and theoretically guaranteed. Extensive experiments on real datasets demonstrate the superiority of our algorithm on both efficacy and efficiency over both macroscopic level (clustering) and microscopic level (structural hole detection) applications.