Due to the non-stationary environment, learning in multi-agent systems is a challenging problem. This paper first introduces a new gradient-based learning algorithm, augmenting the basic gradient ascent approach with policy prediction. We prove that this augmentation results in a stronger notion of convergence than the basic gradient ascent, that is, strategies converge to a Nash equilibrium within a restricted class of iterated games. Motivated by this augmentation, we then propose a new practical multi-agent reinforcement learning (MARL) algorithm exploiting approximate policy prediction. Empirical results show that it converges faster and in a wider variety of situations than state-of-the-art MARL algorithms.

Transfer learning aims at reusing the knowledge in some source tasks to improve the learning of a target task. Many transfer learning methods assume that the source tasks and the target task be related, even though many tasks are not related in reality. However, when two tasks are unrelated, the knowledge extracted from a source task may not help, and even hurt, the performance of a target task. Thus, how to avoid negative transfer and then ensure a "safe transfer" of knowledge is crucial in transfer learning. In this paper, we propose an Adaptive Transfer learning algorithm based on Gaussian Processes (AT-GP), which can be used to adapt the transfer learning schemes by automatically estimating the similarity between a source and a target task. The main contribution of our work is that we propose a new semi-parametric transfer kernel for transfer learning from a Bayesian perspective, and propose to learn the model with respect to the target task, rather than all tasks as in multi-task learning. We can formulate the transfer learning problem as a unified Gaussian Process (GP) model. The adaptive transfer ability of our approach is verified on both synthetic and real-world datasets.

Evolutionary trees of species can be reconstructed by pairwise comparison of their entire genomes. Such a comparison can be quantified by determining the number of events that change the order of genes in a genome. Earlier Erdem and Tillier formulated the pairwise comparison of entire genomes as the problem of planning rearrangement events that transform one genome to the other. We reformulate this problem as a planning problem to extend its applicability to genomes with multiple copies of genes and with unequal gene content, and illustrate its applicability and effectiveness on three real datasets: mitochondrial genomes of Metazoa, chloroplast genomes of Campanulaceae, chloroplast genomes of various land plants and green algae.

We intend to develop auction-like algorithms for the allocation It is often important to coordinate teams of cooperative of complex tasks, similar to SSI auctions for the allocation agents in a distributed manner. We study how to assign of simple tasks. SSI auctions assign simple tasks to tasks to cooperative agents so that the resulting team cost agents in multiple rounds. In each round, each agent bids on is small (that is, team performance is high). Market-based each unassigned task the minimal increase in its agent cost mechanisms are promising distributed task-allocation methods. in case it has to perform this task in addition to all tasks already Robotics researchers have recently studied how to use assigned to it in previous rounds.

Many problems in information extraction, text mining, natural language processing and other fields exhibit the same property: multiple prediction tasks are related in the sense that their outputs (labels) satisfy certain constraints. In this paper, we propose an active learning framework exploiting such relations among tasks. Intuitively, with task outputs coupled by constraints, active learning can utilize not only the uncertainty of the prediction in a single task but also the inconsistency of predictions across tasks. We formalize this idea as a cross-task value of information criteria, in which the reward of a labeling assignment is propagated and measured over all relevant tasks reachable through constraints. A specific example of our framework leads to the cross entropy measure on the predictions of coupled tasks, which generalizes the entropy in the classical single-task uncertain sampling. We conduct experiments on two real-world problems: web information extraction and document classification. Empirical results demonstrate the effectiveness of our framework in actively collecting labeled examples for multiple related tasks.

Human vision is often guided by instinctual commonsense such as proportions and contours. In this paper, we explore how to use the proportion as the key knowledge for designing a privacy algorithm that detects human private parts in a 3D scan dataset. The Analogia Graph is introduced to study the proportion of structures. It is a graph-based representation of the proportion knowledge. The intrinsic human proportions are applied to reduce the search space by an order of magnitude. A feature shape template is constructed to match the model data points using Radial Basis Functions in a non-linear regression and the relative measurements of the height and area factors. The method is tested on 100 datasets from CAESAR database. Two surface rendering methods are studied for data privacy: blurring and transparency. It is found that test subjects normally prefer to have the most possible privacy in both rendering methods. However, the subjects adjusted their privacy measurement to a certain degree as they were informed the context of security.

Graphic design is the process of creating graphics to meet specific commercial needs based on knowledge of layout principles and esthetic concepts. This is usually an iterative trial and error process which requires a lot of time even for expert designers. This expert knowledge can be modelled, represented and used by a computer to perform design activities. This paper describes a novel approach named Gaudii (standing for "Intelligent Automated Graphic Design Generator") which utilizes principles and techniques known from the fields of Evolutionary Computation and Fuzzy Logic to automatically obtain design elements. Experimental results that demonstrate the potential of the proposed approach are presented in the area of poster design.

Although falls are undesirable in humanoid robots, they are also inevitable, especially as robots get deployed in physically interactive human environments. We consider the problem of fall prediction, i.e., to predict if a robot's balance controller can prevent a fall from the current state. A trigger from the fall predictor is used to switch the robot from a balance maintenance mode to a fall control mode. Hence, it is desirable for the fall predictor to signal imminent falls with sufficient lead time before the actual fall, while minimizing false alarms. Analytical techniques and intuitive rules fail to satisfy these competing objectives on a large robot that is subjected to strong disturbances and therefore exhibits complex dynamics. Today effective supervised learning tools are available for finding patterns in high-dimensional data. Our paper contributes a novel approach to engineer fall data such that a supervised learning method can be exploited to achieve reliable prediction. Specifically, we introduce parameters to control the tradeoff between the false positive rate and lead time. Several parameter combinations yield solutions that improve both the false positive rate and the lead time of hand-coded solutions. Learned predictors are decision lists with typical depths of 5-10, in a 16-dimensional feature space. Experiments are carried out in simulation on an Asimo-like robot.

Software defect prediction aims to reduce software testing efforts by guiding testers through the defect-prone sections of software systems. Defect predictors are widely used in organizations to predict defects in order to save time and effort as an alternative to other techniques such as manual code reviews. The application of a defect prediction model in a real-life setting is difficult because it requires software metrics and defect data from past projects to predict the defect-proneness of new projects. It is, on the other hand, very practical because it is easy to apply, can detect defects using less time and reduces the testing effort. We have built a learning-based defect prediction model for a telecommunication company during a period of one year. In this study, we have briefly explained our model, presented its pay-off and described how we have implemented the model in the company. Furthermore, we have compared the performance of our model with that of another testing strategy applied in a pilot project that implemented a new process called Team Software Process (TSP). Our results show that defect predictors can be used as supportive tools during a new process implementation, predict 75% of code defects, and decrease the testing time compared with 25% of the code defects detected through more labor-intensive strategies such as code reviews and formal checklists.

To meet the growing demands of the National Airspace System (NAS) stakeholders and provide the level of service, safety and security needed to sustain future air transport, the Next Generation Air Transportation System (NextGen) concept calls for technologies and systems offering increasing support from automated systems that provide decision-aiding and optimization capabilities. This is an exciting application for some core aspects of Artificial Intelligence research since the automation must be designed to enable the human operators to access and process a myriad of information sources, understand heightened system complexity, and maximize capacity, throughput and fuel savings in the NAS.. This paper introduces an emerging application of techniques from mixed initiative (adjustable autonomy), multi-agent systems, and task scheduling techniques to the air traffic control domain. Consequently, we have created a testbed for investigating the critical challenges in supporting the early design of systems that allow for optimal, context-sensitive function (role) allocation between air traffic controller and automated agents. A pilot study has been conducted with the testbed and preliminary results show a marked qualitative improvement in using dynamic function allocation optimization versus static function allocation.