Replay memory plays an important role in stable learning and fast convergence of deep reinforcement learning algorithms [1] that are methods of approximating a value or a policy function using deep neural networks [2]. The study of replay memory in reinforcement learning started from [3] and played a major role in training reinforcement learning agents to play Atari 2600 games with a Deep Q-Network (DQN) [4]. In addition, replay memory is used in other off-policy reinforcement learning algorithms such as DDPG [5] and ACER [6]. In [7], after analyzing the importance of the data in the replay memory, a probability distribution is assigned to enable efficient learning through prioritization based on the Figure 1: Proposed dual memory structure.

In this paper, we consider sequential online prediction (SOP) for streaming data in the presence of outliers and change points. We propose an INstant TEmporal structure Learning (INTEL) algorithm to address this problem.Our INTEL algorithm is developed based on a full consideration to the duality between online prediction and anomaly detection. We first employ a mixture of weighted GP models (WGPs) to cover the expected possible temporal structures of the data. Then, on the basis of the rich modeling capacity of this WGP mixture, we develop an efficient technique to instantly learn (capture) the temporal structure of the data that follows a regime shift. This instant learning is achieved only by adjusting one hyper-parameter value of the mixture model. A weighted generalization of the product of experts (POE) model is used for fusing predictions yielded from multiple GP models. An outlier is declared once a real observation seriously deviates from the fused prediction. If a certain number of outliers are consecutively declared, then a change point is declared. Extensive experiments are performed using a diverse of real datasets. Results show that the proposed algorithm is significantly better than benchmark methods for SOP in the presence of outliers and change points.

A method of simultaneously optimizing both the structure of neural networks and the connection weights in a single training loop can reduce the enormous computational cost of neural architecture search. We focus on the probabilistic model-based dynamic neural network structure optimization that considers the probability distribution of structure parameters and simultaneously optimizes both the distribution parameters and connection weights based on gradient methods. Since the existing algorithm searches for the structures that only minimize the training loss, this method might find overly complicated structures. In this paper, we propose the introduction of a penalty term to control the model complexity of obtained structures. We formulate a penalty term using the number of weights or units and derive its analytical natural gradient. The proposed method minimizes the objective function injected the penalty term based on the stochastic gradient descent. We apply the proposed method in the unit selection of a fully-connected neural network and the connection selection of a convolutional neural network. The experimental results show that the proposed method can control model complexity while maintaining performance.

We provide a computationally and statistically efficient method for estimating the parameters of a stochastic Gaussian model observed on a spatial grid, which need not be rectangular. Standard methods are plagued by computational intractability, where designing methods that can be implemented for realistically sized problems has been an issue for a long time. This has motivated the use of the Fourier Transform and the Whittle likelihood approximation. The challenge of frequency-domain methods is to determine and account for observational boundary effects, missing data, and the shape of the observed spatial grid. In this paper we address these effects explicitly by proposing a new quasi-likelihood estimator. We prove consistency and asymptotic normality of our estimator, and show that the proposed method solves boundary issues with Whittle estimation for finite samples, yielding parameter estimates with significantly reduced bias and error. We demonstrate the effectiveness of our method for incomplete lattices, in comparison to other recent methods. Finally, we apply our method to estimate the parameters of a Mat\'ern process used to model data from Venus' topography.

Decision trees with binary splits are popularly constructed using Classification and Regression Trees (CART) methodology. For regression models, this approach recursively divides the data into two near-homogenous daughter nodes according to a split point that maximizes the reduction in sum of squares error (the impurity) along a particular variable. This paper aims to study the bias and adaptive properties of regression trees constructed with CART. In doing so, we derive an interesting connection between the bias and the mean decrease in impurity (MDI) measure of variable importance---a tool widely used for model interpretability---defined as the sum of impurity reductions over all non-terminal nodes in the tree. In particular, we show that the size of a terminal subnode for a variable is small when the MDI for that variable is large and that this relationship is exponential---confirming theoretically that decision trees with CART have small bias and are adaptive to signal strength and direction. Finally, we apply these individual tree bounds to tree ensembles and show consistency of Breiman's random forests. The context is surprisingly general and applies to a wide variety of multivariable data generating distributions and regression functions. The main technical tool is an exact characterization of the conditional probability content of the daughter nodes arising from an optimal split, in terms of the partial dependence function and reduction in impurity.

Principles of cognitive economy would require that concepts about objects, properties and relations should be introduced only if they simplify the conceptualisation of a domain. Unexpectedly, classic logic conditionals, specifying structures holding within elements of a formal conceptualisation, do not always satisfy this crucial principle. The paper argues that this requirement is captured by \emph{supervenience}, hereby further identified as a property necessary for compression. The resulting theory suggests an alternative explanation of the empirical experiences observable in Wason's selection tasks, associating human performance with conditionals on the ability of dealing with compression, rather than with logic necessity.

Matching tile games are an extremely popular game genre. Arguably the most popular iteration, Match-3 games, are simple to understand puzzle games, making them great benchmarks for research. In this paper, we propose developing different procedural personas for Match-3 games in order to approximate different human playstyles to create an automated playtesting system. The procedural personas are realized through evolving the utility function for the Monte Carlo Tree Search agent. We compare the performance and results of the evolution agents with the standard Vanilla Monte Carlo Tree Search implementation as well as to a random move-selection agent. We then observe the impacts on both the game's design and the game design process. Lastly, a user study is performed to compare the agents to human play traces.

Games have benchmarked AI methods since of a single game, discovering a few new variations on the inception of the field, with classic board games such existing research topics. The set of · Deckbuilding · Gameplaying · Player Modeling AI problems associated with video games has in recent decades expanded from simply playing games to win, to playing games in particular styles, generating game content, 1 Introduction modeling players etc. Different games pose very different challenges for AI systems, and several different For decades classic board games such as Chess, Checkers, AI challenges can typically be posed by the same and Go have dominated the landscape of AI and game. In this article we analyze the popular collectible games research. Often called the "drosophila of AI" in card game Hearthstone (Blizzard 2014) and describe reference to the drosophila fly's significance in biological a varied set of interesting AI challenges posed by this research, Chess in particular has been the subject game. Collectible card games are relatively understudied of hundreds of academic papers and decades of research in the AI community, despite their popularity and [18]. At the core of many of these approaches is designing the interesting challenges they pose. Analyzing a single algorithms to beat top human players. However, game in-depth in the manner we do here allows us to despite IBM's Deep Blue defeating Garry Kasparov in see the entire field of AI and Games through the lens the 1997 World Chess Championships and DeepMind's AlphaGo defeating Lee Sedol in the 2016 Google Deep-Mind Challenge Match [47], such programs have yet While there is value in designing algorithms to win (e.g.

Users often fail to formulate their complex information needs in a single query. As a consequence, they may need to scan multiple result pages or reformulate their queries, which may be a frustrating experience. Alternatively, systems can improve user satisfaction by proactively asking questions of the users to clarify their information needs. Asking clarifying questions is especially important in conversational systems since they can only return a limited number of (often only one) result(s). In this paper, we formulate the task of asking clarifying questions in open-domain information-seeking conversational systems. To this end, we propose an offline evaluation methodology for the task and collect a dataset, called Qulac, through crowdsourcing. Our dataset is built on top of the TREC Web Track 2009-2012 data and consists of over 10K question-answer pairs for 198 TREC topics with 762 facets. Our experiments on an oracle model demonstrate that asking only one good question leads to over 170% retrieval performance improvement in terms of P@1, which clearly demonstrates the potential impact of the task. We further propose a retrieval framework consisting of three components: question retrieval, question selection, and document retrieval. In particular, our question selection model takes into account the original query and previous question-answer interactions while selecting the next question. Our model significantly outperforms competitive baselines. To foster research in this area, we have made Qulac publicly available.

Graphs are an essential part of many machine learning problems such as analysis of parse trees, social networks, knowledge graphs, transportation systems, and molecular structures. Applying machine learning in these areas typically involves learning the graph structure and the relationship between the nodes of the graph. However, learning the graph structure is often complex, particularly when the graph is cyclic, and the transitions from one node to another are conditioned such as graphs used to represent a finite state machine. To solve this problem, we propose to extend the memory based Neural Turing Machine (NTM) with two novel additions. We allow for transitions between nodes to be influenced by information received from external environments, and we let the NTM learn the context of those transitions. We refer to this extension as the Conditional Neural Turing Machine (CNTM). We show that the CNTM can infer conditional transition graphs by empirically verifiying the model on two data sets: a large set of randomly generated graphs, and a graph modeling the information retrieval process during certain crisis situations. The results show that the CNTM is able to reproduce the paths inside the graph with accuracy ranging from 82,12% for 10 nodes graphs to 65,25% for 100 nodes graphs.