The quality of opponent Artificial Intelligence (AI) in fighting videogames is crucial. Some other game genres can rely on their story or visuals, but fighting games are all about the adversarial experience. In this paper, we will introduce standard behavior algorithms in videogames, such as Finite-State Machines and Behavior Trees, as well as more recent developments, such as Monte-Carlo Tree Search. We will also discuss the existing and potential combinations of these algorithms, and how they might be used in fighting games. Since we are at the financial peak of fighting games, both for casual players and in tournaments, it is important to build and expand on fighting game AI, as it is one of the pillars of this growing market.

Videogames are nowadays one of the biggest entertainment industries in the world. Being part of this industry means competing against lots of other companies and developers, thus, making fanbases of vital importance. They are a group of clients that constantly support your company because your video games are fun. Videogames are most entertaining when the difficulty level is a good match for the player's skill, increasing the player engagement. However, not all players are equally proficient, so some kind of difficulty selection is required. In this paper, we will present Dynamic Difficulty Adjustment (DDA), a recently arising research topic, which aims to develop an automated difficulty selection mechanism that keeps the player engaged and properly challenged, neither bored nor overwhelmed. We will present some recent research addressing this issue, as well as an overview of how to implement it. Satisfactorily solving the DDA problem directly affects the player's experience when playing the game, making it of high interest to any game developer, from independent ones, to 100 billion dollar businesses, because of the potential impacts in player retention and monetization.

To explore heterogeneous information networks (HINs), network representation learning (NRL) is proposed, which represents a network in a low-dimension space. Recently, graph neural networks (GNNs) have drawn a lot of attention which are very expressive for mining a HIN, while they suffer from low efficiency issue. In this paper, we propose a pre-training and fine-tuning framework PF-HIN to capture the features of a HIN. Unlike traditional GNNs that have to train the whole model for each downstream task, PF-HIN only needs to fine-tune the model using the pre-trained parameters and minimal extra task-specific parameters, thus improving the model efficiency and effectiveness. Specifically, in pre-training phase, we first use a ranking-based BFS strategy to form the input node sequence. Then inspired by BERT, we adopt deep bi-directional transformer encoders to train the model, which is a variant of GNN aggregator that is more powerful than traditional deep neural networks like CNN and LSTM. The model is pre-trained based on two tasks, i.e., masked node modeling (MNM) and adjacent node prediction (ANP). Additionally, we leverage factorized embedding parameterization and cross-layer parameter sharing to reduce the parameters. In fine-tuning stage, we choose four benchmark downstream tasks, i.e., link prediction, similarity search, node classification and node clustering. We use node sequence pairs as input for link prediction and similarity search, and a single node sequence as input for node classification and clustering. The experimental results of the above tasks on four real-world datasets verify the advancement of PF-HIN, as it outperforms state-of-the-art alternatives consistently and significantly.

Each Fourth of July for the past five years I've written about AI with the potential to positively impact democratic societies. I return to this question with the hope of shining a light on technology that can strengthen communities, protect privacy and freedoms, or otherwise support the public good. This series is grounded in the principle that artificial intelligence can is capable of not just value extraction, but individual and societal empowerment. While AI solutions often propagate bias, they can also be used to detect that bias. As Dr. Safiya Noble has pointed out, artificial intelligence is one of the critical human rights issues of our lifetimes.

Secretary of State Mike Pompeo seized on a U.N. report confirming Iranian weapons were used to attack Saudi Arabia in September and were part of an arms shipment seized months ago off Yemen's coast; State Department correspondent Rich Edson reports. A fire and an explosion struck a centrifuge production plant above Iran's underground Natanz nuclear enrichment facility early Thursday, analysts said, one of the most-tightly guarded sites in all of the Islamic Republic after earlier acts of sabotage there. The Atomic Energy Organization of Iran sought to downplay the fire, calling it an "incident" that only affected an under-construction "industrial shed," spokesman Behrouz Kamalvandi said. However, both Kamalvandi and Iranian nuclear chief Ali Akbar Salehi rushed after the fire to Natanz, a facility earlier targeted by the Stuxnet computer virus and built underground to withstand enemy airstrikes. The fire threatened to rekindle wider tensions across the Middle East, similar to the escalation in January after a U.S. drone strike killed a top Iranian general in Baghdad and Tehran launched a retaliatory ballistic missile attack targeting American forces in Iraq. While offering no cause for Thursday's blaze, Iran's state-run IRNA news agency published a commentary addressing the possibility of sabotage by enemy nations such as Israel and the U.S. following other recent explosions in the country.

We consider a recently introduced class of network construction problems where edges of a transportation network need to be constructed by a server (construction crew). The server has a constant construction speed which is much lower than its travel speed, so relocation times are negligible with respect to construction times. It is required to find a construction schedule that minimizes a non-decreasing function of the times when various connections of interest become operational. Most problems of this class are strongly NP-hard on general networks, but are often tree-efficient, that is, polynomially solvable on trees. We develop a generic local search heuristic approach and two metaheuristics (Iterated Local Search and Tabu Search) for solving tree-efficient network construction problems on general networks, and explore them computationally. Results of computational experiments indicate that the methods have excellent performance.

As we seek to deploy machine learning models beyond virtual and controlled domains, it is critical to analyze not only the accuracy or the fact that it works most of the time, but if such a model is truly robust and reliable. This paper studies strategies to implement adversary robustly trained algorithms towards guaranteeing safety in machine learning algorithms. We provide a taxonomy to classify adversarial attacks and defenses, formulate the Robust Optimization problem in a min-max setting and divide it into 3 subcategories, namely: Adversarial (re)Training, Regularization Approach, and Certified Defenses. We survey the most recent and important results in adversarial example generation, defense mechanisms with adversarial (re)Training as their main defense against perturbations. We also survey mothods that add regularization terms that change the behavior of the gradient, making it harder for attackers to achieve their objective. Alternatively, we've surveyed methods which formally derive certificates of robustness by exactly solving the optimization problem or by approximations using upper or lower bounds. In addition, we discuss the challenges faced by most of the recent algorithms presenting future research perspectives.

A long-term goal of reinforcement learning agents is to be able to perform tasks in complex real-world scenarios. The use of external information is one way of scaling agents to more complex problems. However, there is a general lack of collaboration or interoperability between different approaches using external information. In this work, we propose a conceptual framework and taxonomy for assisted reinforcement learning, aimed at fostering such collaboration by classifying and comparing various methods that use external information in the learning process. The proposed taxonomy details the relationship between the external information source and the learner agent, highlighting the process of information decomposition, structure, retention, and how it can be used to influence agent learning. As well as reviewing state-of-the-art methods, we identify current streams of reinforcement learning that use external information in order to improve the agent's performance and its decision-making process. These include heuristic reinforcement learning, interactive reinforcement learning, learning from demonstration, transfer learning, and learning from multiple sources, among others. These streams of reinforcement learning operate with the shared objective of scaffolding the learner agent. Lastly, we discuss further possibilities for future work in the field of assisted reinforcement learning systems.

As demand for computer software continually increases, software scope and complexity become higher than ever. The software industry is in real need of accurate estimates of the project under development. Software development effort estimation is one of the main processes in software project management. However, overestimation and underestimation may cause the software industry loses. This study determines which technique has better effort prediction accuracy and propose combined techniques that could provide better estimates. Eight different ensemble models to estimate effort with Ensemble Models were compared with each other base on the predictive accuracy on the Mean Absolute Residual (MAR) criterion and statistical tests. The results have indicated that the proposed ensemble models, besides delivering high efficiency in contrast to its counterparts, and produces the best responses for software project effort estimation. Therefore, the proposed ensemble models in this study will help the project managers working with development quality software.

The world came together to build 5G. Now the next-generation wireless technology is pulling the world apart. The latest version of the 5G technical specifications, expected Friday, adds features for connecting autonomous cars, intelligent factories, and internet-of-things devices to crazy-fast 5G networks. The blueprints reflect a global effort to develop the technology, with contributions from more than a dozen companies from Europe, the US, and Asia. And yet, 5G is also pulling nations apart--with the US and China anchoring the tug-of-war.