Browsing by Author "Bilbao, Miren Nekane"
Now showing 1 - 13 of 13
Results Per Page
Sort Options
Item Big Data for transportation and mobility: recent advances, trends and challenges: Recent advances, trends and challenges(2018-10-04) Torre-Bastida, Ana I.; Del Ser, Javier; Laña, Ibai; Ilardia, Maitena; Bilbao, Miren Nekane; Campos-Cordobes, Sergio; Tecnalia Research & Innovation; HPA; IA; LABORATORIO DE TRANSFORMACIÓN URBANA; SMART_TRANSPORTBig Data is an emerging paradigm and has currently become a strong attractor of global interest, specially within the transportation industry. The combination of disruptive technologies and new concepts such as the Smart City upgrades the transport data life cycle. In this context, Big Data is considered as a new pledge for the transportation industry to effectively manage all data this sector required for providing safer, cleaner and more efficient transport means, as well as for users to personalize their transport experience. However, Big Data comes along with its own set of technological challenges, stemming from the multiple and heterogeneous transportation/mobility application scenarios. In this survey we analyze the latest research efforts revolving on Big Data for the transportation and mobility industry, its applications, baselines scenarios, fields and use case such as routing, planning, infrastructure monitoring, network design, among others. This analysis will be done strictly from the Big Data perspective, focusing on those contributions gravitating on techniques, tools and methods for modeling, processing, analyzing and visualizing transport and mobility Big Data. From the literature review a set of trends and challenges is extracted so as to provide researchers with an insightful outlook on the field of transport and mobility.Item Dandelion-encoded harmony search heuristics for opportunistic traffic offloading in synthetically modeled mobile networks(Springer Verlag, 2016) Perfecto, Cristina; Bilbao, Miren Nekane; Del Ser, Javier; Ferro, Armando; Salcedo-Sanz, Sancho; Geem, Zong Woo; Kim, Joong Hoon; IAThe high data volumes being managed by and transferred through mobile networks in the last few years are the main rationale for the upsurge of research aimed at finding efficient technical means to offload exceeding traffic to alternative communication infrastructures with higher transmission bandwidths. This idea is solidly buttressed by the proliferation of short-range wireless communication technologies (e.g.mobile devices with multiple radio interfaces), which can be conceived as available opportunistic hotspots to which the operator can reroute exceeding network traffic depending on the contractual clauses of the owner at hand. Furthermore, by offloading to such hotspots a higher effective coverage can be attained by those operators providing both mobile and fixed telecommunication services. In this context, the operator must decide if data generated by its users will be sent over conventional 4G+/4G/3G communication links, or if they will instead be offloaded to nearby opportunistic networks assuming a contractual cost penalty. Mathematically speaking, this problem can be formulated as a spanning tree optimization subject to cost-performance criteria and coverage constraints. This paper will elaborate on the efficient solving of this optimization paradigm by means of the Harmony Search meta-heuristic algorithm and the so-called Dandelion solution encoding, the latter allowing for the use of conventional meta-heuristic operators maximally preserving the locality of tree representations. The manuscript will discuss the obtained simulation results over different synthetically modeled setups of the underlying communication scenario and contractual clauses of the users.Item A discrete water cycle algorithm for solving the symmetric and asymmetric traveling salesman problem(2018-10) Osaba, Eneko; Del Ser, Javier; Sadollah, Ali; Bilbao, Miren Nekane; Camacho, David; Quantum; IAThe water cycle algorithm (WCA) is a nature-inspired meta-heuristic recently contributed to the community in 2012, which finds its motivation in the natural surface runoff phase in water cycle process and on how streams and rivers flow into the sea. This method has been so far successfully applied to many engineering applications, spread over a wide variety of application fields. In this paper an enhanced discrete version of the WCA (coined as DWCA) is proposed for solving the Symmetric and Asymmetric Traveling Salesman Problem. Aimed at proving that the developed approach is a promising approximation method for solving this family of optimization problems, the designed solver has been tested over 33 problem datasets, comparing the obtained outcomes with the ones got by six different algorithmic counterparts from the related literature: genetic algorithm, island-based genetic algorithm, evolutionary simulated annealing, bat algorithm, firefly algorithm and imperialist competitive algorithm. Furthermore, the statistical significance of the performance gaps found in this benchmark is validated based on the results from non-parametric tests, not only in terms of optimality but also in regards to convergence speed. We conclude that the proposed DWCA approach outperforms – with statistical significance – any other optimization technique in the benchmark in terms of both computation metrics.Item Evolving Spiking Neural Networks for online learning over drifting data streams(2018-12) Lobo, Jesus L.; Laña, Ibai; Del Ser, Javier; Bilbao, Miren Nekane; Kasabov, Nikola; IANowadays huge volumes of data are produced in the form of fast streams, which are further affected by non-stationary phenomena. The resulting lack of stationarity in the distribution of the produced data calls for efficient and scalable algorithms for online analysis capable of adapting to such changes (concept drift). The online learning field has lately turned its focus on this challenging scenario, by designing incremental learning algorithms that avoid becoming obsolete after a concept drift occurs. Despite the noted activity in the literature, a need for new efficient and scalable algorithms that adapt to the drift still prevails as a research topic deserving further effort. Surprisingly, Spiking Neural Networks, one of the major exponents of the third generation of artificial neural networks, have not been thoroughly studied as an online learning approach, even though they are naturally suited to easily and quickly adapting to changing environments. This work covers this research gap by adapting Spiking Neural Networks to meet the processing requirements that online learning scenarios impose. In particular the work focuses on limiting the size of the neuron repository and making the most of this limited size by resorting to data reduction techniques. Experiments with synthetic and real data sets are discussed, leading to the empirically validated assertion that, by virtue of a tailored exploitation of the neuron repository, Spiking Neural Networks adapt better to drifts, obtaining higher accuracy scores than naive versions of Spiking Neural Networks for online learning environments.Item A feature selection method for author identification in interactive communications based on supervised learning and language typicality(2016-11-01) Villar-Rodriguez, Esther; Del Ser, Javier; Bilbao, Miren Nekane; Salcedo-Sanz, Sancho; Tecnalia Research & Innovation; Quantum; IAAuthorship attribution, conceived as the identification of the origin of a text between different authors, has been a very active area of research in the scientific community mainly supported by advances in Natural Language Processing (NLP), machine learning and Computational Intelligence. This paradigm has been mostly addressed from a literary perspective, aiming at identifying the stylometric features and writeprints which unequivocally typify the writer patterns and allow their unique identification. On the other hand, the upsurge of social networking platforms and interactive messaging have undoubtedly made the anonymous expression of feelings, the sharing of experiences and social relationships much easier than in other traditional communication media. Unfortunately, the popularity of such communities and the virtual identification of their users deploy a rich substrate for cybercrimes against unsuspecting victims and other forms of illegal uses of social networks that call for the activity tracing of accounts. In the context of one-to-one communications this manuscript postulates the identification of the sender of a message as a useful approach to detect impersonation attacks in interactive communication scenarios. In particular this work proposes to select linguistic features extracted from messages via NLP techniques by means of a novel feature selection algorithm based on the dissociation between essential traits of the sender and receiver influences. The performance and computational efficiency of different supervised learning models when incorporating the proposed feature selection method is shown to be promising with real SMS data in terms of identification accuracy, and paves the way towards future research lines focused on applying the concept of language typicality in the discourse analysis field.Item A harmony search approach for the selective pick-up and delivery problem with delayed drop-off(Springer Verlag, 2016) Del Ser, Javier; Bilbao, Miren Nekane; Perfecto, Cristina; Salcedo-Sanz, Sancho; Geem, Zong Woo; Kim, Joong Hoon; IAIn the last years freight transportation has undergone a sharp increase in the scales of its underlying processes and protocols mainly due to the ever-growing community of users and the increasing number of on-line shopping stores. Furthermore, when dealing with the last stage of the shipping chain an additional component of complexity enters the picture as a result of the fixed availability of the destination of the good to be delivered. As such, business opening hours and daily work schedules often clash with the delivery times programmed by couriers along their routes. In case of conflict, the courier must come to an arrangement with the destination of the package to be delivered or, alternatively, drop it off at a local depot to let the destination pick it up at his/her time convenience. In this context this paper will formulate a variant of the so-called courier problem under economic profitability criteria including the cost penalty derived from the delayed drop-off. In this context, if the courier delivers the package to its intended destination before its associated deadline, he is paid a reward. However, if he misses to deliver in time, the courier may still deliver it at the destination depending on its availability or, alternatively, drop it off at the local depot assuming a certain cost. The manuscript will formulate the mathematical optimization problem that models this logistics process and solve it efficiently by means of the Harmony Search algorithm. A simulation benchmark will be discussed to validate the solutions provided by this meta-heuristic solver and to compare its performance to other algorithmic counterparts.Item Hybridizing Cartesian Genetic Programming and Harmony Search for adaptive feature construction in supervised learning problems(2016) Elola, Andoni; Del Ser, Javier; Bilbao, Miren Nekane; Perfecto, Cristina; Alexandre, Enrique; Salcedo-Sanz, Sancho; IAThe advent of the so-called Big Data paradigm has motivated a flurry of research aimed at enhancing machine learning models by following very diverse approaches. In this context this work focuses on the automatic construction of features in supervised learning problems, which differs from the conventional selection of features in that new characteristics with enhanced predictive power are inferred from the original dataset. In particular this manuscript proposes a new iterative feature construction approach based on a self-learning meta-heuristic algorithm (Harmony Search) and a solution encoding strategy (correspondingly, Cartesian Genetic Programming) suited to represent combinations of features by means of constant-length solution vectors. The proposed feature construction algorithm, coined as Adaptive Cartesian Harmony Search (ACHS), incorporates modifications that allow exploiting the estimated predictive importance of intermediate solutions and, ultimately, attaining better convergence rate in its iterative learning procedure. The performance of the proposed ACHS scheme is assessed and compared to that rendered by the state of the art in a toy example and three practical use cases from the literature. The excellent performance figures obtained in these problems shed light on the widespread applicability of the proposed scheme to supervised learning with legacy datasets composed by already refined characteristics.Item Let nature decide its nature: On the design of collaborative hyperheuristics for decentralized ephemeral environments: On the design of collaborative hyperheuristics for decentralized ephemeral environments(2018-11) Martinez, Aritz; Osaba, Eneko; Bilbao, Miren Nekane; Ser, Javier Del; Quantum; IAThe research community has traditionally aimed at the derivation and development of metaheuristic solvers, suited to deal with problems of very diverse characteristics. Unfortunately, it is often the case that new metaheuristic techniques are presented and assessed in a reduced set of cases, mostly due to the lack of computational resources to undertake extensive performance studies over a sufficiently diverse set of optimization benchmarks. This manuscript explores how ephemeral environments could be exploited to efficiently construct metaheuristic algorithms by virtue of a collaborative, distributed nature-inspired hyperheuristic framework specifically designed to be deployed over unreliable, uncoordinated computation nodes. To this end, the designed framework defines two types of nodes (trackers and peers, similarly to peer-to-peer networks), both reacting resiliently to unexpected disconnections of nodes disregarding their type. Peer nodes exchange their populations (i.e. constructed algorithms) asynchronously, so that local optima are avoided at every peer thanks to the contribution by other nodes. Furthermore, the overall platform is fully scalable, allowing its users to implement and share newly derived operators and fitness functions so as to enrich the diversity and universality of the heuristic algorithms found by the framework. Results obtained from in-lab experiments with a reduced number of nodes are discussed to shed light on the evolution of the best solution of the framework with the number of connected peers and the tolerance of the network to node disconnections.Item A novel Grouping Coral Reefs Optimization algorithm for optimal mobile network deployment problems under electromagnetic pollution and capacity control criteria(2016-08-15) Salcedo-Sanz, Sancho; Garcia-Diaz, Pilar; Del Ser, Javier; Bilbao, Miren Nekane; Portilla-Figueras, Jose Antonio; IAThis paper proposes a novel optimization algorithm for grouping problems, the Grouping Coral Reefs Optimization algorithm, and describes its application to a Mobile Network Deployment Problem (MNDP) under four optimization criteria. These criteria include economical cost and coverage, and also electromagnetic pollution control and capacity constraints imposed at the base stations controllers, which are novel in this study. The Coral Reefs Optimization algorithm (CRO) is a recently-proposed bio-inspired approach for optimization, based on the simulation of the processes that occur in coral reefs, including reproduction, fight for space or depredation. This paper presents a grouping version of the CRO, which has not previously evaluated before. Grouping meta-heuristics are characterized by variable-length encoding solutions, and have been successfully applied to a number of different optimization and assignment problems. The GCRO proposed is a novel contribution to the intelligent systems field, which is able to improve results obtained by two alternative grouping algorithms such as grouping genetic algorithms and grouping Harmony Search. The performance of the proposed GCRO and the algorithms for comparison has been tested with real data in a case study of a MNDP in Alcalá de Henares, Madrid, Spain.Item On the Application of Multi-objective Harmony Search Heuristics to the Predictive Deployment of Firefighting Aircrafts: a Realistic Case Study: A realistic case study(2015) Bilbao, Miren Nekane; Del Ser, Javier; Salcedo-Sanz, Sancho; Casanova-Mateo, Carlos; IAThis manuscript focuses on the increasing frequency and scales of worldwide wildfires and the need for enhancing the effectiveness of firefighting resources. The scope is focused on optimally deploying firefighting aircrafts on aerodromes and airports existing over an area based on fire risk predictions. This scenario is formulated as a capacity-constrained multi-objective optimisation problem where the utility of the deployed resources with respect to fire forest risk predictions is to be maximised, and expenditures associated with the reallocation of aircrafts must be minimised. This formulation is further complemented by including the impact of the distance from the wildfire to water sources in the firefighting utility function. To efficiently tackle this problem a multi-objective harmony search solver is designed and tested in synthetically generated and real scenarios for the Iberian Peninsula. The results obtained pave the way towards the utilisation of this tool by decision makers when outlining their firefighting logistics.Item On the heritability of dandelion-encoded harmony search heuristics for tree optimization problems(IEEE, 2015-09-24) Perfecto, Cristina; Bilbao, Miren Nekane; Del Ser, Javier; Ferro, Armando; IATree based optimization problems stand for those paradigms where solutions can be arranged within a tree-like graph whose nodes represent the optimization variables of the problem at hand and their interconnecting edges topological and/or hierarchical relationships between such variables. In this context, a research line of increasing interest during the last decade focuses on the derivation of intelligent solution encoding strategies capable of 1) capturing all topological constraints of this particular class of graphs; and 2) preserving their connectivity properties when they undergo combination/mutation operations within approximative evolutionary solvers. This manuscript takes a step over the state of the art by shedding light on the heri-tability properties of the Dandelion tree encoding approach under avant-garde stochastically-controlled evolutionary operators. In particular we elaborate on the topological heritability of the so-called Harmony Memory Considering Rate (HMCR) exploitative operator of the Harmony Search algorithm, a population-based meta-heuristic algorithm that has so far shown to outperform other evolutionary schemes in a wide range of optimization scenarios. Results from extensive Monte Carlo simulations are discussed in terms of the preserved structural properties of the newly produced solutions with respect to the initial Dandelion-encoded population.Item On the post-hoc explainability of deep echo state networks for time series forecasting, image and video classification(2022-07) Barredo Arrieta, Alejandro; Gil-Lopez, Sergio; Laña, Ibai; Bilbao, Miren Nekane; Del Ser, Javier; Tecnalia Research & Innovation; IASince their inception, learning techniques under the reservoir computing paradigm have shown a great modeling capability for recurrent systems without the computing overheads required for other approaches, specially deep neural networks. Among them, different flavors of echo state networks have attracted many stares through time, mainly due to the simplicity and computational efficiency of their learning algorithm. However, these advantages do not compensate for the fact that echo state networks remain as black-box models whose decisions cannot be easily explained to the general audience. This issue is even more involved for multi-layered (also referred to as deep) echo state networks, whose more complex hierarchical structure hinders even further the explainability of their internals to users without expertise in machine learning or even computer science. This lack of explainability can jeopardize the widespread adoption of these models in certain domains where accountability and understandability of machine learning models is a must (e.g., medical diagnosis, social politics). This work addresses this issue by conducting an explainability study of echo state networks when applied to learning tasks with time series, image and video data. Among these tasks, we stress on the latter one (video classification) which, to the best of our knowledge, has never been tackled before with echo state networks in the related literature. Specifically, the study proposes three different techniques capable of eliciting understandable information about the knowledge grasped by these recurrent models, namely potential memory, temporal patterns and pixel absence effect. Potential memory addresses questions related to the effect of the reservoir size in the capability of the model to store temporal information, whereas temporal patterns unveil the recurrent relationships captured by the model over time. Finally, pixel absence effect attempts at evaluating the effect of the absence of a given pixel when the echo state network model is used for image and video classification. The benefits of the proposed suite of techniques are showcased over three different domains of applicability: time series modeling, image and, for the first time in the related literature, video classification. The obtained results reveal that the proposed techniques not only allow for an informed understanding of the way these models work, but also serve as diagnostic tools capable of detecting issues inherited from data (e.g., presence of hidden bias).Item A Probabilistic Sample Matchmaking Strategy for Imbalanced Data Streams with Concept Drift(2017) L. Lobo, Jesus; Del Ser, Javier; Bilbao, Miren Nekane; Laña, Ibai; Salcedo-Sanz, Sancho; IAIn the last decade the interest in adaptive models for non-stationary environments has gained momentum within the research community due to an increasing number of application scenarios generating non-stationary data streams. In this context the literature has been specially rich in terms of ensemble techniques, which in their majority have focused on taking advantage of past information in the form of already trained predictive models and other alternatives alike. This manuscript elaborates on a rather different approach, which hinges on extracting the essential predictive information of past trained models and determining therefrom the best candidates (intelligent sample matchmaking) for training the predictive model of the current data batch. This novel perspective is of inherent utility for data streams characterized by short-length unbalanced data batches, situation where the so-called trade-off between plasticity and stability must be carefully met. The approach is evaluated on a synthetic data set that simulates a non-stationary environment with recurrently changing concept drift. The proposed approach is shown to perform competitively when adapting to a sudden and recurrent change with respect to the state of the art, but without storing all the past trained models and by lessening its computational complexity in terms of model evaluations. These promising results motivate future research aimed at validating the proposed strategy on other scenarios under concept drift, such as those characterized by semi-supervised data streams.