In this paper we introduce the problem of planning for perception of a target position. Given a sensing target, the robot has to move to a goal position from where the target can be perceived. Our algorithm minimizes the overall path cost as a function of both motion and perception costs, given an initial robot position and a sensing target. We contribute a heuristic search method, PA*, that efficiently searches for an optimal path. We prove the proposed heuristic is admissible, and introduce a new goal state stopping condition.
In this paper, we analyze domain bias in automated text-based personality prediction, and proposes a novel method to correct domain bias. The proposed approach is very general since it requires neither retraining a personality prediction system using examples from a new domain, nor any knowledge of the original training data used to develop the system. We conduct several experiments to evaluate the effectiveness of the method, and the findings indicate a significant improvement of prediction accuracy.
Explicit communication planning is an increasing necessity for agent systems. Furthermore, there has also been a renewed interest in using classical planning to perform this planning in addition to physical actions in a goal-directed way. Existing approaches, however, are not applicable to a broad spectrum of domains. We present several generic pre-processing strategies and adaptations of the Fast-Forward (FF) planner that we compare in two different domains in terms of time complexity.
In this paper a simulation-based approach to finding optimal defender strategy in multi-act Security Games (SG) played on a graph is proposed. The method employs the Upper Confidence Bounds applied to Trees (UCT) algorithm which relies on massive simulations of possible game scenarios. Three different variants of the algorithm are presented and compared with each other as well as against the Mixed Integer Linear Program (MILP) exact solution in terms of computational efficiency and memory requirements. Experimental evaluation shows that the method has a few times lower memory demands and is faster than MILP approach in majority of test cases while preserving quality of the resulting mixed strategies.
We propose a new strategic model of negotiation, called Boolean negotiation games. Our model is inspired by Boolean games and the alternating offers model of bargaining. It offers a computationally grounded model for studying properties of negotiation protocols in a qualitative setting. Boolean negotiation games can yield agreements that are more beneficial than stable solutions (Nash equilibria) of the underlying Boolean game.
Shahrzad Gholami, Bryan Wilder, Matthew Brown, Dana Thomas, Nicole Sintov, Milind Tambe
1750 - 1751
Security agencies including the US Coast Guard, the Federal Air Marshal Service and the Los Angeles Airport police are several major domains that have been deploying Stackelberg security games and related algorithms to protect against a single adversary or multiple, independent adversaries strategically. However, there are a variety of real-world security domains where adversaries may benefit from colluding in their actions against the defender. Given the potential negative effect of these collusive actions, the defender has an incentive to break up collusion by playing off the self-interest of individual adversaries. This paper deals with problem of collusive security games for rational and bounded rational adversaries. The theoretical results verified with human subject experiments showed that behavior model which optimizes against bounded rational adversaries provides demonstrably better performing defender strategies against human subjects.
Communities typically capture homophily as people of the same community share many common features. This paper is motivated by the problem of community detection in social networks, as it can help improve our understanding of the network topology. Given the selfish nature of humans to align with like-minded people, we employ game theoretic models and algorithms to detect communities in this paper. Specifically, we employ coordination games to represent interactions between individuals in a social network. We provide a novel and scalable two phased algorithm NashOverlap to compute an accurate overlapping community structure in the given network. We evaluate our algorithm against the best existing methods for community detection and show that our algorithm improves significantly on benchmark networks with respect to standard normalised mutual information measure.
Sarvar Abdullaev, Peter McBurney, Katarzyna Musial
1754 - 1755
Options constitute integral part of modern financial trades, and are priced according to the risk associated with buying or selling certain asset in future. Financial literature mostly concentrates on risk-neutral methods of pricing options such as Black-Scholes model. However, it is an emerging field in option pricing theory to use trading agents with utility functions to determine the option's potential payoff for the agent. In this paper, we use one of such methodologies developed by Othman and Sandholm to design portfolio-holding agents that are endowed with popular option portfolios such as bullish spread, butterfly spread, straddle, etc to price options. Agents use their portfolios to evaluate how buying or selling certain option would change their current payoff structure, and form their orders based on this information. We also simulate these agents in a multi-unit direct double auction. The emerging prices are compared to risk-neutral prices under different market conditions. Through an appropriate endowment of option portfolios to agents, we can also mimic market conditions where the population of agents are bearish, bullish, neutral or non-neutral in their beliefs.
We propose a novel Hierarchical CBC model (HCBC) based on Formal Concept Analysis (FCA). Firstly, Concept Lattice (CL), the hierarchical and conceptual structure in FCA, is adopted to represent cases. Thus a novel dynamic weight model is proposed from CL to measure similarities between cases and concepts. Then the similarity metric is applied to retrieve the top-K similar concepts which are used to vote for adaptive solutions for new cases by majority voting in case adaption. Experiments show our model shows good performance in terms of accuracy and outperforms the other classification methods.
This paper proposes a novel SAPNet model that incorporates a stochastic area pooling (SAP) method with a generic stacked T-shaped CNN architecture. In our SAP method, pooling area is randomly transformed and max pooling operation is then conducted on such areas, which are no longer regular identical fixed upright squares. It can be viewed as feature-level augmentation, substantially reducing model parameters while keeping generalization ability of CNN almost unchanged. Furthermore, we present a generic CNN architecture that structurally resembles three stacked T-shaped cubes. In such architecture, the number of kernels in convolutional layer preceding any pooling layer is doubled and all learnable weight layers are combined with batch normalization and dropout with a small ratio. Finally, on CIFAR-10, CIFAR-100, MNIST, and SVHN datasets, the experimental results show that our SAPNet requires fewer parameters than regular CNN models and still achieves superior recognition performances for all the four benchmarks.
The capability of determining the right sequence of physical actions to achieve a given task is essential for AI that interacts with the physical world. The great difficulty in developing this capability has two main causes: (1) the world is continuous and therefore the action space is infinite, (2) due to noisy perception, we do not know the exact physical properties of our environment and therefore cannot precisely simulate the consequences of a physical action.
In this paper we define a realistic physical action selection problem that has many features common to these kind of problems, the minigolf hole-in-one problem: given a two-dimensional minigolf-like obstacle course, a ball and a hole, determine a single shot that hits the ball into the hole. We assume gravity as well as noisy perception of the environment. We present a method that solves this problem similar to how humans are approaching these problems, by using qualitative reasoning and mental simulation, combined with sampling of actions in the real environment and adjusting the internal knowledge based on observing the actual outcome of sampled actions. We evaluate our method using difficult minigolf levels that require the ball to bounce at several objects in order to hit the hole and compare with existing methods.
Leonor Becerra-Bonache, Hendrik Blockeel, María Galván, François Jacquenet
1764 - 1765
In the past, research on learning language models mainly used syntactic information during the learning process but in recent years, researchers began to also use semantic information. This paper presents such an approach where the input of our learning algorithm is a dataset of pairs made up of sentences and the contexts in which they are produced. The system we present is based on inductive logic programming techniques that aim to learn a mapping between n-grams and a semantic representation of their associated meaning. Experiments have shown that we can learn such a mapping that made it possible later to generate relevant descriptions of images or learn the meaning of words without any linguistic resource.
Y. Gatsoulis, M.O. Mehmood, V.G. Dimitrova, D.R. Magee, B. Sage-Vallier, P. Thiaudiere, J. Valdes, A.G. Cohn
1769 - 1774
The transport network in many countries relies on extended portions which run underground in tunnels. As tunnels age, repairs are required to prevent dangerous collapses. However repairs are expensive and will affect the operational efficiency of the tunnel. We present a decision support system (DSS) based on supervised machine learning methods that learns to predict the risk factor and the resulting repair urgency in the tunnel maintenance planning of a European national rail operator. The data on which the prototype has been built consists of 47 tunnels of varying lengths. For each tunnel, periodic survey inspection data is available for multiple years, as well as other data such as the method of construction of the tunnel. Expert annotations are also available for each 10m tunnel segment for each survey as to the degree of repair urgency which are used for both training and model evaluation. We show that good predictive power can be obtained and discuss the relative merits of a number of learning methods.
As the world becomes increasingly digitally readable through a variety of sensors, digital services will play a key role in advising and supporting people towards a variety of goals. In this paper, we present a personalized wellness system that leverages techniques from cognitive science and machine learning to improve a user's well-being by suggesting daily micro-goals (e.g., “bring a healthy snack to work”), and by enabling social sharing of individual achievements. Specifically, we propose a method for estimating a user's likelihood of successfully completing a given micro-goal (“ONE”) and study the correlation between ONEs and users' actions to improve their chances of reaching their wellness objectives.
Chris A.B. Baker, Sarvapali Ramchurn, W.T. Luke Teacy, Nicholas R. Jennings
1777 - 1782
The coordination of multiple Unmanned Aerial Vehicles (UAVs) to carry out aerial surveys is a major challenge for emergency responders. In particular, UAVs have to fly over kilometre-scale areas while trying to discover casualties as quickly as possible. To aid in this process, it is desirable to exploit the increasing availability of data about a disaster from sources such as crowd reports, satellite remote sensing, or manned reconnaissance. In particular, such information can be a valuable resource to drive the planning of UAV flight paths over a space in order to discover people who are in danger. However challenges of computational tractability remain when planning over the very large action spaces that result. To overcome these, we introduce the survivor discovery problem and present as our solution, the first example of a continuous factored coordinated Monte Carlo tree search algorithm. Our evaluation against state of the art benchmarks show that our algorithm, Co-CMCTS, is able to localise more casualties faster than standard approaches by 7% or more on simulations with real-world data.
Since malaria is prevalent in less developed and more remote areas in which public health resources are often scarce, targeted intervention is essential in allocating resources for effective malaria control. To effectively support targeted intervention, predictive models must be not only accurate but they must also have high temporal and spatial resolution to help determine when and where to intervene. In this paper we take the first essential step towards a system to support targeted intervention in Thailand by developing a high resolution prediction model through the combination of Bayes nets and ARIMA. Bayes nets and ARIMA have complementary strengths, with the Bayes nets better able to represent the effect of environmental variables and ARIMA better able to capture the characteristics of the time series of malaria cases. Leveraging these complementary strengths, we develop an ensemble predictor from the two that has significantly better accuracy that either predictor alone. We build and test the models with data from Tha Song Yang district in northern Thailand, creating village-level models with weekly temporal resolution.
Armon Toubman, Jan Joris Roessingh, Pieter Spronck, Aske Plaat, Jaap van den Herik
1791 - 1796
Adaptive behaviour for computer generated forces enriches training simulations with appropriate challenge levels. For adequate insight into the range of possible behaviour, the adaptation has to take place in a rapid fashion. Ideally, each new behaviour model should remain readable by (and thereby under the control of) human experts. Although various attempts have been made at creating adaptive behaviour, current solutions require large numbers of simulations. Moreover, usability by end users has been of subordinate interest, as is compliance with doctrine and ethics. In this work, we present a machine learning method that enables fast behaviour adaptation, while keeping the behaviour models in a human-readable format. We demonstrate the effectiveness of the proposed method in beyond-visual-range air combat simulations.
Aldy Gunawan, Hoong Chuin Lau, Pradeep Varakantham, Wenjie Wang
1797 - 1802
Many conference mobile apps today lack the intelligent feature to automatically generates optimal schedules based on delegates' preferences. This entails two major challenges: (a) identifying preferences of users; and (b) given the preferences, generating a schedule that optimizes his preferences. In this paper, we specifically focus on academic conferences, where users are prompted to input their preferred keywords. Our key contribution is an integrated conference scheduling agent that automatically recognizes user preferences based on keywords, provides a list of recommended talks and optimizes user schedule based on these preferences. To demonstrate the utility of our integrated conference scheduling agent, we first demonstrated the app in the International Conference on Autonomous Agents and Multi-Agent Systems (AAMAS 2015) and conducted a survey to collect some data, which are used to verify the results presented in this paper. It is able to provide well calibrated results with respect to precision, accuracy and recall. We also tested the app in the 2015 WI-IAT International Conference (Singapore). The android and web-based apps have been demonstrated and deployed in AAMAS 2016 (Singapore) with positive responses from the users.
Martin Gjoreski, Hristijan Gjoreski, Mitja Luštrek, Matjaž Gams
1803 - 1804
In this paper we propose a method for continuous stress monitoring using data provided by a commercial wrist device equipped with common physiological sensors and an accelerometer. The method consists of three machine-learning components: a laboratory stress-detector that detects short-term stress every 2 minutes; an activity recognizer that continuously recognizes user's activity and thus provides context information; and a context-based stress detector that first aggregates the predictions of the laboratory detector, and then exploits the user's context in order to provide the final decision in a 20 minute interval. The method was trained on 21 subjects in a laboratory setting and tested on 5 subjects in a real-life setting. The accuracy on 55 days of real-life data was 92%. The method is currently being implemented as a smartphone application, which will be demonstrated at the conference.
Artificial Intelligence techniques are increasingly being used to develop smart training applications for professionals in various domains. This paper presents an intelligent training system that enables professionals in the public domain to practice their aggression de-escalation skills. The system is one of the main products of the STRESS project, an interdisciplinary research project involving partners from academia, industry and society. The system makes use of a variety of AI-related techniques, including simulation, virtual agents, sensor fusion, model-based analysis and adaptive support. A preliminary evaluation of the system has been conducted with two groups of potential end users, namely tram conductors and police academy students.
Teena Hassan, Dominik Seuss, Johannes Wollenberg, Jens Garbas, Ute Schmid
1812 - 1817
In many domains of computer vision, such as medical imaging and facial image analysis, it is necessary to combine shape (geometric) and appearance (texture) information. In this paper, we describe a method for combining geometric and texture-based evidence for facial actions within a Kalman filter framework. The geometric evidence is provided by a face alignment method. The texture-based evidence is provided by a set of Support Vector Machines (SVM) for various Action Units (AU). The proposed method is a practical solution to the problem of fusing categorical probabilities within a Kalman filter based state estimation framework. A first performance evaluation on upper face AUs demonstrates the practical applicability of the proposed fusion method. The method is applicable to arbitrary imaging domains, apart from facial action estimation.
This paper describes e-Tourism2.0, a web-based recommendation and planning system for tourism activities that takes into account the preferences that define the travel style of the user. e-Tourism2.0 features a recommender system with access to various web services in order to obtain updated information about locations, monuments, opening hours, or transportation modes. The planning system of e-Tourism2.0 models the taste and travel style preferences of the user and creates a planning problem which is later solved by a planner, returning a personalized plan (agenda) for the tourist. e-Tourism2.0 contributes with a special module that calculates the recommendable duration of a visit for a user and the modeling of preferences into a planning problem.