This paper reviews the field of Game AI, which not only deals with creating agents that can play a certain game, but also with areas as diverse as creating game content automatically, game analytics, or player modelling. While Game AI was for a long time not very well recognized by the larger scientific community, it has established itself as a research area for developing and testing the most advanced forms of AI algorithms and articles covering advances in mastering video games such as StarCraft 2 and Quake III appear in the most prestigious journals. Because of the growth of the field, a single review cannot cover it completely. Therefore, we put a focus on important recent developments, including that advances in Game AI are starting to be extended to areas outside of games, such as robotics or the synthesis of chemicals. In this article, we review the algorithms and methods that have paved the way for these breakthroughs, report on the other important areas of Game AI research, and also point out exciting directions for the future of Game AI.

Graph convolutional networks (GCNs) are a family of neural network models that perform inference on graph data by interleaving vertex-wise operations and message-passing exchanges across nodes. Concerning the latter, two key questions arise: (i) how to design a differentiable exchange protocol (e.g., a 1-hop Laplacian smoothing in the original GCN), and (ii) how to characterize the trade-off in complexity with respect to the local updates. In this paper, we show that state-of-the-art results can be achieved by adapting the number of communication steps independently at every node. In particular, we endow each node with a halting unit (inspired by Graves' adaptive computation time) that after every exchange decides whether to continue communicating or not. We show that the proposed adaptive propagation GCN (AP-GCN) achieves superior or similar results to the best proposed models so far on a number of benchmarks, while requiring a small overhead in terms of additional parameters. We also investigate a regularization term to enforce an explicit trade-off between communication and accuracy. The code for the AP-GCN experiments is released as an open-source library.

Generative moment matching networks (GMMNs) are introduced as dependence models for the joint innovation distribution of multivariate time series (MTS). Following the popular copula-GARCH approach for modeling dependent MTS data, a framework allowing us to take an alternative GMMN-GARCH approach is presented. First, ARMA-GARCH models are utilized to capture the serial dependence within each univariate marginal time series. Second, if the number of marginal time series is large, principal component analysis (PCA) is used as a dimension-reduction step. Last, the remaining cross-sectional dependence is modeled via a GMMN, our main contribution. GMMNs are highly flexible and easy to simulate from, which is a major advantage over the copula-GARCH approach. Applications involving yield curve modeling and the analysis of foreign exchange rate returns are presented to demonstrate the utility of our approach, especially in terms of producing better empirical predictive distributions and making better probabilistic forecasts. All results are reproducible with the demo GMMN_MTS_paper of the R package gnn.

Trustworthy AI is a critical issue in machine learning where, in addition to training a model that is accurate, one must consider both fair and robust training in the presence of data bias and poisoning. However, the existing model fairness techniques mistakenly view poisoned data as an additional bias, resulting in severe performance degradation. To fix this problem, we propose FR-Train, which holistically performs fair and robust model training. We provide a mutual information-based interpretation of an existing adversarial training-based fairness-only method, and apply this idea to architect an additional discriminator that can identify poisoned data using a clean validation set and reduce its influence. In our experiments, FR-Train shows almost no decrease in fairness and accuracy in the presence of data poisoning by both mitigating the bias and defending against poisoning. We also demonstrate how to construct clean validation sets using crowdsourcing, and release new benchmark datasets.

We propose the SupRB learning system, a new Pittsburgh-style learning classifier system (LCS) for supervised learning on multi-dimensional continuous decision problems. SupRB learns an approximation of a quality function from examples (consisting of situations, choices and associated qualities) and is then able to make an optimal choice as well as predict the quality of a choice in a given situation. One area of application for SupRB is parametrization of industrial machinery. In this field, acceptance of the recommendations of machine learning systems is highly reliant on operators' trust. While an essential and much-researched ingredient for that trust is prediction quality, it seems that this alone is not enough. At least as important is a human-understandable explanation of the reasoning behind a recommendation. While many state-of-the-art methods such as artificial neural networks fall short of this, LCSs such as SupRB provide human-readable rules that can be understood very easily. The prevalent LCSs are not directly applicable to this problem as they lack support for continuous choices. This paper lays the foundations for SupRB and shows its general applicability on a simplified model of an additive manufacturing problem.

Fully data-driven, deep learning-based models are usually designed as language-independent and have been shown to be successful for many natural language processing tasks. However, when the studied language is low-resourced and the amount of training data is insufficient, these models can benefit from the integration of natural language grammar-based information. We propose two approaches to dependency parsing especially for languages with restricted amount of training data. Our first approach combines a state-of-the-art deep learning-based parser with a rule-based approach and the second one incorporates morphological information into the parser. In the rule-based approach, the parsing decisions made by the rules are encoded and concatenated with the vector representations of the input words as additional information to the deep network. The morphology-based approach proposes different methods to include the morphological structure of words into the parser network. Experiments are conducted on the IMST-UD Treebank and the results suggest that integration of explicit knowledge about the target language to a neural parser through a rule-based parsing system and morphological analysis leads to more accurate annotations and hence, increases the parsing performance in terms of attachment scores. The proposed methods are developed for Turkish, but can be adapted to other languages as well.

Algorithmic fairness has been studied mostly in a static setting where the implicit assumptions are that the frequencies of historically made decisions do not impact the problem structure in subsequent future. However, for example, the capability to pay back a loan for people in a certain group might depend on historically how frequently that group has been approved loan applications. If banks keep rejecting loan applications to people in a disadvantaged group, it could create a feedback loop and further damage the chance of getting loans for people in that group. This challenge has been noted in several recent works but is under-explored in a more generic sequential learning setting. In this paper, we formulate this delayed and long-term impact of actions within the context of multi-armed bandits (MAB). We generalize the classical bandit setting to encode the dependency of this action "bias" due to the history of the learning. Our goal is to learn to maximize the collected utilities over time while satisfying fairness constraints imposed over arms' utilities, which again depend on the decision they have received. We propose an algorithm that achieves a regret of $\tilde{\mathcal{O}}(KT^{2/3})$ and show a matching regret lower bound of $\Omega(KT^{2/3})$, where $K$ is the number of arms and $T$ denotes the learning horizon. Our results complement the bandit literature by adding techniques to deal with actions with long-term impacts and have implications in designing fair algorithms.

Federated learning, as an emerging distributed training model of neural networks without collecting raw data, has attracted widespread attention. However, almost all existing researches of federated learning only consider protecting the privacy of clients, but not preventing model iterates and final model parameters from leaking to untrusted clients and external attackers. In this paper, we present the first bilateral privacy-preserving federated learning scheme, which protects not only the raw training data of clients, but also model iterates during the training phase as well as final model parameters. Specifically, we present an efficient privacy-preserving technique to mask or encrypt the global model, which not only allows clients to train over the noisy global model, but also ensures only the server can obtain the exact updated model. Detailed security analysis shows that clients can access neither model iterates nor the final global model; meanwhile, the server cannot obtain raw training data of clients from additional information used for recovering the exact updated model. Finally, extensive experiments demonstrate the proposed scheme has comparable model accuracy with traditional federated learning without bringing much extra communication overhead.

We consider the post-training quantization problem, which discretizes the weights of pre-trained deep neural networks without re-training the model. We propose multipoint quantization, a quantization method that approximates a full-precision weight vector using a linear combination of multiple vectors of low-bit numbers; this is in contrast to typical quantization methods that approximate each weight using a single low precision number. Computationally, we construct the multipoint quantization with an efficient greedy selection procedure, and adaptively decides the number of low precision points on each quantized weight vector based on the error of its output. This allows us to achieve higher precision levels for important weights that greatly influence the outputs, yielding an 'effect of mixed precision' but without physical mixed precision implementations (which requires specialized hardware accelerators). Empirically, our method can be implemented by common operands, bringing almost no memory and computation overhead. We show that our method outperforms a range of state-of-the-art methods on ImageNet classification and it can be generalized to more challenging tasks like PASCAL VOC object detection.

With the increasing use of online social networks as a source of news and information, the propensity for a rumor to disseminate widely and quickly poses a great concern, especially in disaster situations where users do not have enough time to fact-check posts before making the informed decision to react to a post that appears to be credible. In this study, we explore the propagation pattern of rumors on Twitter by exploring the dynamics of microscopic-level misinformation spread, based on the latent message and user interaction attributes. We perform supervised learning for feature selection and prediction. Experimental results with real-world data sets give the models' prediction accuracy at about 90\% for the diffusion of both True and False topics. Our findings confirm that rumor cascades run deeper and that rumor masked as news, and messages that incite fear, will diffuse faster than other messages. We show that the models for True and False message propagation differ significantly, both in the prediction parameters and in the message features that govern the diffusion. Finally, we show that the diffusion pattern is an important metric in identifying the credibility of a tweet.