How do you create an intelligent player for a game? Artificial intelligence offers a variety of ways to program intelligence into computer opponents. In this article, we'll show how it works, using intelligent heuristics and a web-based game that you can try yourself. Artificial intelligence is becoming an increasingly important topic in the field of computer science. While advancements in machine learning continue to break records in areas including image recognition, voice recognition, translation, and natural language processing, many additional branches of AI continue to advance as well. One of the earliest applications of AI is in the area of game development. Specifically, artificial intelligence is often used to create opponent players in games. Early forms of AI players in games often consisted of traditional board games, such as chess, checkers, backgammon, and tic-tac-toe. Games of this type provide a fully observable and deterministic view at any point in the state of the game. This allows an AI player the ability to analyze all possible moves from both the human player and the AI player itself, thus determining the best likely move to take at any given time. AI players in video games have since expanded to a much broader range of gaming categories, where the best move or course of action is not always crystal clear. These include games that often utilize random events or actions, in addition to hidden views of the game or of the opponent's actions.

Neural Architecture Search (NAS) was first proposed to achieve state-of-the-art performance through the discovery of new architecture patterns, without human intervention. An over-reliance on expert knowledge in the search space design has however led to increased performance (local optima) without significant architectural breakthroughs, thus preventing truly novel solutions from being reached. In this work we propose 1) to cast NAS as a problem of finding the optimal network generator and 2) a new, hierarchical and graph-based search space capable of representing an extremely large variety of network types, yet only requiring few continuous hyper-parameters. This greatly reduces the dimensionality of the problem, enabling the effective use of Bayesian Optimisation as a search strategy. At the same time, we expand the range of valid architectures, motivating a multi-objective learning approach. We demonstrate the effectiveness of our strategy on six benchmark datasets and show that our search space generates extremely lightweight yet highly competitive models illustrating the benefits of a NAS approach that optimises over network generator selection.

Recent works using deep learning to solve the Traveling Salesman Problem (TSP) have focused on learning construction heuristics. Such approaches find TSP solutions of good quality but require additional procedures such as beam search and sampling to improve solutions and achieve state-of-the-art performance. However, few studies have focused on improvement heuristics, where a given solution is improved until reaching a near-optimal one. In this work, we propose to learn a local search heuristic based on 2-opt operators via deep reinforcement learning. We propose a policy gradient algorithm to learn a stochastic policy that selects 2-opt operations given a current solution. Moreover, we introduce a policy neural network that leverages a pointing attention mechanism, which unlike previous works, can be easily extended to more general k-opt moves. Our results show that the learned policies can improve even over random initial solutions and approach near-optimal solutions at a faster rate than previous state-of-the-art deep learning methods.

We present the Goal Uncertain Stochastic Shortest Path (GUSSP) problem -- a general framework to model path planning and decision making in stochastic environments with goal uncertainty. The framework extends the stochastic shortest path (SSP) model to dynamic environments in which it is impossible to determine the exact goal states ahead of plan execution. GUSSPs introduce flexibility in goal specification by allowing a belief over possible goal configurations. The unique observations at potential goals helps the agent identify the true goal during plan execution. The partial observability is restricted to goals, facilitating the reduction to an SSP with a modified state space. We formally define a GUSSP and discuss its theoretical properties. We then propose an admissible heuristic that reduces the planning time using FLARES -- a start-of-the-art probabilistic planner. We also propose a determinization approach for solving this class of problems. Finally, we present empirical results on a search and rescue mobile robot and three other problem domains in simulation.

In this article, we introduce PackingSolver, a new solver for two-dimensional two- and three-staged guillotine Packing Problems. It relies on a simple yet powerful anytime tree search algorithm called Memory Bounded A* (MBA*). This algorithm was first introduced in libralesso2020 for the 2018 ROADEF/EURO challenge glass cutting problem (https://www.roadef.org/challenge/2018/en/index.php), for which it had been ranked first during the final phase. In this article, we generalize it for a large variety of Cutting and Packing problems. The resulting program can tackle two-dimensional Bin Packing, Multiple Knapsack, and Strip Packing Problems, with two- or three-staged exact or non-exact guillotine cuts, the orientation of the first cut being imposed or not, and with or without item rotation. Despite its simplicity and genericity, computational experiments show that this approach is competitive compared to the other dedicated algorithms from the literature. It even returns state-of-the-art solutions on several variants. The combination of efficiency, ability to provide good solutions fast, simplicity and versatility makes it particularly suited for industrial applications, which require quickly developing algorithms implementing several business-specific constraints.

Metaheuristic search strategies have proven their effectiveness against man-made solutions in various contexts. They are generally effective in local search area exploitation, and their overall performance is largely impacted by the balance between exploration and exploitation. Recent developments in parallel local search explore methods to take advantage of the efficient local exploitation of searches and reach impressive results. This however restricts the scaling potential to nodes within a private, trusted computer cluster. In this research we propose a novel blockchain protocol that allows parallel local search to scale to untrusted and anonymous computational nodes. The protocol introduces publicly verifiable performance evaluation of the local optima reported by each node, creating a competitive environment between the local searches. That is strengthened with economical stimuli for producing good solutions, that provide coordination between the nodes, as every node tries to explore different sections of the search space to beat their competition.

In this article, we present the anytime tree search algorithm we designed for the 2018 ROADEF/EURO challenge glass cutting problem proposed by the French company Saint-Gobain. The resulting program was ranked first among 64 participants. Its key components are: a new search algorithm called Memory Bounded A* (MBA*) with guide functions, a symmetry breaking strategy, and a pseudo-dominance rule. We perform a comprehensive study of these components showing that each of them contributes to the algorithm global performances. In addition, we designed a second tree search algorithm fully based on the pseudo-dominance rule and dedicated to some of the challenge instances with strong precedence constraints. On these instances, it finds the best-known solutions very quickly.

Decades of research have been invested in making computer programs for playing games such as Chess and Go. This paper focuses on a new game, Tetris Link, a board game that is still lacking any scientific analysis. Tetris Link has a large branching factor, hampering a traditional heuristic planning approach. We explore heuristic planning and two other approaches: Reinforcement Learning, Monte Carlo tree search. We document our approach and report on their relative performance in a tournament. Curiously, the heuristic approach is stronger than the planning/learning approaches. However, experienced human players easily win the majority of the matches against the heuristic planning AIs. We, therefore, surmise that Tetris Link is more difficult than expected. We offer our findings to the community as a challenge to improve upon.

Conformance checking encompasses a body of process mining techniques which aim to find and describe the differences between a process model capturing the expected process behavior and a corresponding event log recording the observed behavior. Alignments are an established technique to compute the distance between a trace in the event log and the closest execution trace of a corresponding process model. Given a cost function, an alignment is optimal when it contains the least number of mismatches between a log trace and a model trace. Determining optimal alignments, however, is computationally expensive, especially in light of the growing size and complexity of event logs from practice, which can easily exceed one million events with traces of several hundred activities. A common limitation of existing alignment techniques is the inability to exploit repetitions in the log. By exploiting a specific form of sequential pattern in traces, namely tandem repeats, we propose a novel technique that uses pre- and post-processing steps to compress the length of a trace and recomputes the alignment cost while guaranteeing that the cost result never under-approximates the optimal cost. In an extensive empirical evaluation with 50 real-life model-log pairs and against five state-of-the-art alignment techniques, we show that the proposed compression approach systematically outperforms the baselines by up to an order of magnitude in the presence of traces with repetitions, and that the cost over-approximation, when it occurs, is negligible.

We propose to incorporate neural architecture search (NAS) into general-purpose multi-task learning (GP-MTL). Existing NAS methods typically define different search spaces according to different tasks. In order to adapt to different task combinations (i.e., task sets), we disentangle the GP-MTL networks into single-task backbones (optionally encode the task priors), and a hierarchical and layerwise features sharing/fusing scheme across them. This enables us to design a novel and general task-agnostic search space, which inserts cross-task edges (i.e., feature fusion connections) into fixed single-task network backbones. Moreover, we also propose a novel single-shot gradient-based search algorithm that closes the performance gap between the searched architectures and the final evaluation architecture. This is realized with a minimum entropy regularization on the architecture weights during the search phase, which makes the architecture weights converge to near-discrete values and therefore achieves a single model. As a result, our searched model can be directly used for evaluation without (re-)training from scratch. We perform extensive experiments using different single-task backbones on various task sets, demonstrating the promising performance obtained by exploiting the hierarchical and layerwise features, as well as the desirable generalizability to different i) task sets and ii) single-task backbones. The code of our paper is available at https://github.com/bhpfelix/MTLNAS.