Reinforcement learning of pedagogical policies in adaptive and intelligent educational systems

In an adaptive and intelligent educational system (AIES), the process of learning pedagogical policies according the students needs fits as a Reinforcement Learning (RL) problem. Previous works have demonstrated that a great amount of experience is needed in order for the system to learn to teach properly, so applying RL to the AIES from scratch is unfeasible. Other works have previously demonstrated in a theoretical way that seeding the AIES with an initial value function learned with simulated students reduce the experience required to learn an accurate pedagogical policy. In this paper we present empirical results demonstrating that a value function learned with simulated students can provide the AIES with a very accurate initial pedagogical policy. The evaluation is based on the interaction of more than 70 Computer Science undergraduate students, and demonstrates that an efficient and useful guide through the contents of the educational system is obtained.     

Constructive learning neural network applied to identification and control of a fuel-ethanol fermentation process

In the present work, a constructive learning algorithm was employed to design a near-optimal one-hidden layer neural network structure that best approximates the dynamic behavior of a bioprocess. The method determines not only a proper number of hidden neurons but also the particular shape of the activation function for each node. Here, the projection pursuit technique was applied in association with the optimization of the solvability condition, giving rise to a more efficient and accurate computational learning algorithm. As each activation function of a hidden neuron is defined according to the peculiarities of each approximation problem, better rates of convergence are achieved, guiding to parsimonious neural network architectures. The proposed constructive learning algorithm was successfully applied to identify a MIMO bioprocess, providing a multivariable model that was able to describe the complex process dynamics, even in long-range horizon predictions. The resulting identification model was considered as part of a model-based predictive control strategy, producing high-quality performance in closed-loop experiments.     

A simple robust control for global asymptotic position stabilization of underactuated surface vessels

Global asymptotic stabilization of underactuated surface vessels is generally achieved only by designing complicated controllers. This paper proposes a very simple control law that globally asymptotically stabilizes the position of underactuated surface vessel to a desired constant location and its velocities to zero. The proposed controller is independent of velocity signals and is robust to model parameters. It neither includes an observer nor an adaptive/sliding‐mode law. Controller development and stability analysis rely on a novel Lyapunov function and LaSalle's theorem. Furthermore, by extending the proposed control strategy, a saturated control law is also obtained ensuring the semiglobal asymptotic stability of position error system. Effectiveness of the proposed control schemes is demonstrated by simulation examples. Copyright © 2017 John Wiley & Sons, Ltd.     

Supporting Unified Distributed Management and Autonomic Decisions: Design,Implementation and Deployment

Nowadays, the prevailing use of networks based on traditional centralized management systems reflects on a fast increase of the management costs. The growth in the number of network equipments and services reinforces the need to distribute the management responsibilities throughout the network devices. In this approach, each device executes common network management functionalities, being part of the overall network management platform. In this paper, we present a Unified Distributed Network Management (UDNM) framework that provides a unified (wired and wireless) management network solution, where further different network services can take part of this infrastructure, e.g., flow monitoring, accurate routing decisions, distributed policies dissemination, etc. This framework is divided in two main components: (A) Situation awareness, which sets up initial information through bootstrapping, discovery, fault-management process and exchange of management information; (B) Autonomic Decision System (ADS) that performs distributed decisions in the network with incomplete information. We deploy the UDNM framework in a testbed which involves two cities (\(\approx\)250 km between), different standards (IEEE 802.3, IEEE 802.11 and IEEE 802.16e) and network technologies, such as, wired virtual grid, wireless ad-hoc gateways, ad-hoc mobile access devices. The UDNM framework integrates management functionalities into the managed devices, proving to be a lightweight and easy-respond framework. The performance analysis shows that the UDNM framework is feasible to unify devices management functionalities and to take accurate decisions on top of a real network.     

Increasing selective pressure towards the best compromise in evolutionary multiobjective optimization: The extended NOSGA method

Most current approaches in the evolutionary multiobjective optimization literature concentrate on adapting an evolutionary algorithm to generate an approximation of the Pareto frontier. However, finding this set does not solve the problem. The decision-maker still has to choose the best compromise solution out of that set. Here, we introduce a new characterization of the best compromise solution of a multiobjective optimization problem. By using a relational system of preferences based on a multicriteria decision aid way of thinking, and an outranked-based dominance generalization, we derive some necessary and sufficient conditions which describe satisfactory approximations to the best compromise. Such conditions define a lexicographic minimum of a bi-objective optimization problem, which is a map of the original one. The NOSGA-II method is a NSGA-II inspired efficient way of solving the resulting mapped problem.     

Sensitivity of SAR data to post-fire forest regrowth in Mediterranean and boreal forests

Disturbed forests may need decades to reach a mature stage and optically-based vegetation indices are usually poorly suited for monitoring purposes due to the rapid saturation of the signal with increasing canopy cover. Spaceborne synthetic aperture radar (SAR) data provide an alternate monitoring approach since the backscattered microwave energy is sensitive to the vegetation structure. Images from two regions in Spain and Alaska were used to analyze SAR metrics (cross-polarized backscatter and co-polarized interferometric coherence) from regrowing forests previously affected by fire. TerraSAR-X X-band backscatter showed the lowest sensitivity to forest regrowth, with the average backscatter increasing by 1-2 dB between the most recent fire scar and the unburned forest. Increased sensitivity (around 3-4 dB) was observed for C-band Envisat Advanced Synthetic Aperture (ASAR) backscatter. The Advanced Land Observing Satellite (ALOS) Phased Array-type L-band Synthetic Aperture Radar (PALSAR) L-band backscatter presented the highest dynamic range from unburned to recently burned forests (approximately 8 dB). The interferometric coherence showed low sensitivity to forest regrowth at all SAR frequencies. For Mediterranean forests, five phases of forest regrowth were discerned whereas for boreal forest, up to four different regrowth phases could be discerned with L-band SAR data. In comparison, the Normalized Difference Vegetation Index (NDVI) provided reliable differentiation only for the most recent development stages. The results obtained were consistent in both environments.     

Exoskeleton: Curve network abstraction for 3D shapes

In this paper, we introduce the concept of an exoskeleton as a new abstraction of arbitrary shapes that succinctly conveys both the perceptual and the geometric structure of a 3D model. We extract exoskeletons via a principled framework that combines segmentation and shape approximation. Our method starts from a segmentation of the shape into perceptually relevant parts and then constructs the exoskeleton using a novel extension of the Variational Shape Approximation method. Benefits of the exoskeleton abstraction to graphics applications such as simplification and chartification are presented.     

A Graph‐based Approach to Continuous Line Illustrations with Variable Levels of Detail

This paper introduces a method for automatically generating continuous line illustrations, drawings consisting of a single line, from a given input image. Our approach begins by inferring a graph from a set of edges extracted from the image in question and obtaining a path that traverses through all edges of the said graph. The resulting path is then subjected to a series of post‐processing operations to transform it into a continuous line drawing. Moreover, our approach allows us to manipulate the amount of detail portrayed in our line illustrations, which is particularly useful for simplifying the overall illustration while still retaining its most significant features. We also present several experimental results to demonstrate that our approach can automatically synthesize continuous line illustrations comparable to those of some contemporary artists.