Crazyhouse is a chess variant that incorporates all of the classical chess rules, but allows users to drop pieces captured from the opponent as a normal move. Until 2018, all competitive computer engines for this board game made use of an alpha-beta pruning algorithm with a hand-crafted evaluation function for each position. Previous machine learning-based algorithms for just regular chess, such as NeuroChess and Giraffe, took hand-crafted evaluation features as input rather than a raw board representation. More recent projects, such as AlphaZero, reached massive success but required massive computational resources in order to reach its final strength. This paper describes the development of SixtyFour, an engine designed to compete in the chess variant of Crazyhouse with limited hardware. This specific variant poses a multitude of significant challenges due to its large branching factor, state-space complexity, and the multiple move types a player can make. We propose the novel creation of a neural network-based evaluation function for Crazyhouse. More importantly, we evaluate the effectiveness of an ensemble model, which allows the training time and datasets to be easily distributed on regular CPU hardware commodity. Early versions of the network have attained a playing level comparable to a strong amateur on online servers.

"Our AI takes about 20 moves, most of the time solving it in the minimum number of steps," Baldi says. "Right there, you can see the strategy is different, so my best guess is that the AI's form of reasoning is completely different from a human's." The ultimate goal of projects such as this one is to build the next generation of AI systems, Baldi says. Whether they know it or not, artificial intelligence touches people every day through apps such as Siri and Alexa and recommendation engines working behind the scenes of their favorite online services. "But these systems are not really intelligent; they're brittle, and you can easily break or fool them," Baldi says.

We consider a co-variate shift problem where one has access to several marginally different training datasets for the same learning problem and a small validation set which possibly differs from all the individual training distributions. This co-variate shift is caused, in part, due to unobserved features in the datasets. The objective, then, is to find the best mixture distribution over the training datasets (with only observed features) such that training a learning algorithm using this mixture has the best validation performance. Our proposed algorithm, ${\sf Mix\&Match}$, combines stochastic gradient descent (SGD) with optimistic tree search and model re-use (evolving partially trained models with samples from different mixture distributions) over the space of mixtures, for this task. We prove simple regret guarantees for our algorithm with respect to recovering the optimal mixture, given a total budget of SGD evaluations. Finally, we validate our algorithm on two real-world datasets.

In this paper, we introduce the Double Traveling Salesman Problem with Partial Last-In-First-Out Loading Constraints (DTSPPL), a pickup-and-delivery single-vehicle routing problem where all pickup operations must be performed before any delivery one because the pickup and delivery areas are geographically separated. The vehicle collects items in the pickup area and loads them into its container, a horizontal stack. After performing all pickup operations, the vehicle begins delivering the items in the delivery area. Loading and unloading operations must obey a partial Last-In-First-Out (LIFO) policy, i.e., a version of the LIFO policy that may be violated within a given reloading depth. The objective of the DTSPPL is to minimize the total cost, which involves the total distance traveled by the vehicle and the number of reloaded items due to violations of the standard LIFO policy. We formally describe the DTSPPL by means of two Integer Linear Programming (ILP) formulations, and propose a heuristic algorithm based on the Biased Random-Key Genetic Algorithm (BRKGA) to find high-quality solutions. The performance of the proposed solution approaches is assessed over a broad set of instances. Computational results have shown that both ILP formulations were able to solve only the smaller instances, whereas the BRKGA obtained better solutions for almost all instances, requiring shorter computational time.

With the wealth of information produced by social networks, smartphones, medical or financial applications, speculations have been raised about the sensitivity of such data in terms of users' personal privacy and data security. To address the above issues, Federated Learning (FL) has been recently proposed as a means to leave data and computational resources distributed over a large number of nodes (clients) where a central coordinating server aggregates only locally computed updates without knowing the original data. In this work, we extend the FL framework by pushing forward the state the art in the field on several dimensions: (i) unlike the original FedAvg approach relying solely on single criteria (i.e., local dataset size), a suite of domain- and client-specific criteria constitute the basis to compute each local client's contribution, (ii) the multi-criteria contribution of each device is computed in a prioritized fashion by leveraging a priority-aware aggregation operator used in the field of information retrieval, and (iii) a mechanism is proposed for online-adjustment of the aggregation operator parameters via a local search strategy with backtracking. Extensive experiments on a publicly available dataset indicate the merits of the proposed approach compared to standard FedAvg baseline.

In recent years, a growing number of researchers have explored the use of robotic arms or dexterous hands to solve a variety of everyday tasks. While many of them have successfully tackled simple tasks, such as grasping or basic manipulation, complex tasks that involve multiple steps and precise/strategic movements have so far proved harder to address. A team of researchers at the Chinese University of Hong Kong and Tencent AI Lab has recently developed a deep learning-based approach to solve a Rubik's Cube using a multi-fingered dexterous hand. Their approach, presented in a paper pre-published on arXiv, allows a dexterous hand to solve more advanced in-hand manipulation tasks, such as the renowned Rubik's Cube puzzle. A Rubik's Cube is a plastic cube covered in multi-colored squares that can be shifted into different positions.

The influence of deep learning is continuously expanding across different domains, and its new applications are ubiquitous. The question of neural network design thus increases in importance, as traditional empirical approaches are reaching their limits. Manual design of network architectures from scratch relies heavily on trial and error, while using existing pretrained models can introduce redundancies or vulnerabilities. Automated neural architecture design is able to overcome these problems, but the most successful algorithms operate on significantly constrained design spaces, assuming the target network to consist of identical repeating blocks. We propose a probabilistic representation of a neural network structure under the assumption of independence between layer types. The probability matrix (prototype) can describe general feedforward architectures and is equivalent to the population of models, while being simple to interpret and analyze. We construct an architecture search algorithm, inspired by the estimation of distribution algorithms, to take advantage of this representation. The probability matrix is tuned towards generating high-performance models by repeatedly sampling the architectures and evaluating the corresponding networks. Our algorithm is shown to discover models which are competitive with those produced by existing architecture search methods, both in accuracy and computational costs, despite the conceptual simplicity and the comparatively limited scope of achievable designs.

Being able to model correlations between labels is considered crucial in multi-label classification. Rule-based models enable to expose such dependencies, e.g., implications, subsumptions, or exclusions, in an interpretable and human-comprehensible manner. Albeit the number of possible label combinations increases exponentially with the number of available labels, it has been shown that rules with multiple labels in their heads, which are a natural form to model local label dependencies, can be induced efficiently by exploiting certain properties of rule evaluation measures and pruning the label search space accordingly. However, experiments have revealed that multi-label heads are unlikely to be learned by existing methods due to their restrictiveness. To overcome this limitation, we propose a plug-in approach that relaxes the search space pruning used by existing methods in order to introduce a bias towards larger multi-label heads resulting in more expressive rules. We further demonstrate the effectiveness of our approach empirically and show that it does not come with drawbacks in terms of training time or predictive performance.

Multi-modal journey planning, which allows multiple types of transport within a single trip, is becoming increasingly popular, due to a strong practical interest and an increasing availability of data. In real life, transport networks feature uncertainty. Yet, most approaches assume a deterministic environment, making plans more prone to failures such as missed connections and major delays in the arrival. This paper presents an approach to computing optimal contingent plans in multi-modal journey planning. The problem is modeled as a search in an and/or state space. We describe search enhancements used on top of the AO* algorithm. Enhancements include admissible heuristics, multiple types of pruning that preserve the completeness and the optimality, and a hybrid search approach with a deterministic and a nondeterministic search. We demonstrate an NP-hardness result, with the hardness stemming from the dynamically changing distributions of the travel time random variables. We perform a detailed empirical analysis on realistic transport networks from cities such as Montpellier, Rome and Dublin. The results demonstrate the effectiveness of our algorithmic contributions, and the benefits of contingent plans as compared to standard sequential plans, when the arrival and departure times of buses are characterized by uncertainty.

Anthony D. Rhodes Portland State University Abstract -- This paper presents two novel approaches to solving the classic board game mastermind, including a variant of simulated annealing (SA) and a technique we term maximum expected reduction in consistency (MERC). In addition, we compare search results for these algorithms to two baseline search methods: a random, uninformed search and the method of minimizing maximum query partition sets as originally developed by both Donald Knuth [3] and Peter Norvig [4]. I. INTRODUCTION Mastermind is a popular code-breaking two player game originally invented in the 1970s. The gameplay closely resembles the antecedent pen and paper game called "Bulls and Cows", which dates back at least a century. Mastermind consists of three components: a decoding board which includes a dozen or so rows of holes for query pegs, in addition to smaller holes for key pegs; the board also contains a space for the placement of the master code provided by the code-maker.