The development of autonomous agents that perform tasks with the same dexterity as performed by humans is one of the challenges of artificial intelligence and robotics. This motivates the research on intelligent agents, since the agent must choose the best action in a dynamic environment in order to maximise the final score. In this context, the present paper introduces a novel algorithm for Qualitative Case-Based Reasoning and Learning (QCBRL), which is a case-based reasoning system that uses qualitative spatial representations to retrieve and reuse cases by means of relations between objects in the environment. Combined with reinforcement learning, QCBRL allows the agent to learn new qualitative cases at runtime, without assuming a pre-processing step. In order to avoid cases that do not lead to the maximum performance, QCBRL executes case-base maintenance, excluding these cases and obtaining new (more suitable) ones. Experimental evaluation of QCBRL was conducted in a simulated robot-soccer environment, in a real humanoid-robot environment and on simple tasks in two distinct gridworld domains. Results show that QCBRL outperforms traditional RL methods. As a result of running QCBRL in autonomous soccer matches, the robots performed a higher average number of goals than those obtained when using pure numerical models. In the gridworlds considered, the agent was able to learn optimal and safety policies.

This paper presents a new type of evolutionary algorithm (EA) based on the concept of “meme”, where the individuals forming the population are represented by semantic networks and the fitness measure is defined as a function of the represented knowledge. Our work can be classified as a novel memetic algorithm (MA), given that (1) it is the units of culture, or information, that are undergoing variation, transmission, and selection, very close to the original sense of memetics as it was introduced by Dawkins; and (2) this is different from existing MA, where the idea of memetics has been utilized as a means of local refinement by individual learning after classical global sampling of EA. The individual pieces of information are represented as simple semantic networks that are directed graphs of concepts and binary relations, going through variation by memetic versions of operators such as crossover and mutation, which utilize knowledge from commonsense knowledge bases. In evaluating this introductory work, as an interesting fitness measure, we focus on using the structure mapping theory of analogical reasoning from psychology to evolve pieces of information that are analogous to a given base information. Considering other possible fitness measures, the proposed representation and algorithm can serve as a computational tool for modeling memetic theories of knowledge, such as evolutionary epistemology and cultural selection theory.

We propose an efficient method, built on the popular Bag of Features approach, that obtains robust multiclass pixel-level object segmentation of an image in less than 500ms, with results comparable or better than most state of the art methods. We introduce the Integral Linear Classifier (ILC), that can readily obtain the classification score for any image sub-window with only 6 additions and 1 product by fusing the accumulation and classification steps in a single operation. In order to design a method as efficient as possible, our building blocks are carefully selected from the quickest in the state of the art. More precisely, we evaluate the performance of three popular local descriptors, that can be very efficiently computed using integral images, and two fast quantization methods: the Hierarchical K-Means, and the Extremely Randomized Forest. Finally, we explore the utility of adding spatial bins to the Bag of Features histograms and that of cascade classifiers to improve the obtained segmentation. Our method is compared to the state of the art in the difficult Graz-02 and PASCAL 2007 Segmentation Challenge datasets.

Selecting which algorithms should be used by a mobile robot computer vision system is a decision that is usually made a priori by the system developer, based on past experience and intuition, not systematically taking into account information that can be found in the images and in the visual process itself to learn which algorithm should be used, in execution time. This paper presents a method that uses Reinforcement Learning to decide which algorithm should be used to recognize objects seen by a mobile robot in an indoor environment, based on simple attributes extracted on-line from the images, such as mean intensity and intensity deviation. Two state-of-the-art object recognition algorithms can be selected: the constellation method proposed by Lowe together with its interest point detector and descriptor, the Scale-Invariant Feature Transform and a bag of features approach. A set of empirical evaluations was conducted using a household mobile robots image database, and results obtained shows that the approach adopted here is very promising.

This paper presents a vision-based approach for mobile robot localization. The environmental model is topological. The new approach uses a constellation of different types of affine covariant regions to characterize a place. This type of representation permits a reliable and distinctive environment modeling. The performance of the proposed approach is evaluated using a database of panoramic images from different rooms. Additionally, we compare different combinations of complementary feature region detectors to find the one that achieves the best results. Our experimental results show promising results for this new localization method. Additionally, similarly to what happens with single detectors, different combinations exhibit different strengths and weaknesses depending on the situation, suggesting that a context-aware method to combine the different detectors would improve the localization results.

In this article, we build upon a multiagent architecture for landmark-based navigation in unknown environments. In this architecture, each of the agents in the navigation system has a bidding function that is controlled by a set of parameters. We show here the good results obtained by an evolutionary approach that tunes the parameter set values for two navigation tasks. © 2005 Wiley Periodicals, Inc. Int J Int Syst 20: 523–539, 2005.

This paper is an attempt at providing a fuzzy set formalization of case-based reasoning and decision. Learning aspects are not considered here. The proposed approach assumes a principle stating that “the more similar are the problem description attributes, the more similar are the outcome attributes.” A weaker form of this principle concluding only on the graded possibility of the similarity of the outcome attributes, is also considered. These two forms of the case-based reasoning principle are modelled in terms of fuzzy rules. Then an approximate reasoning machinery taking advantage of this principle enables us to apply the information stored in the memory of previous cases to the current problem. A particular instance of case-based reasoning, named case-based decision, is especially investigated. A logical formalization of the basic case-based reasoning inference is also proposed. Extensions of the proposed approach in order to handle imprecise or fuzzy descriptions or to manage more general forms of the principle underlying case-based reasoning are briefly discussed in the conclusion. © 1998 John Wiley & Sons, Inc.

In dealing with representing knowledge under uncertainty there is a sustained tendency to increase flexibility in order to avoid problems of inconsistency in the knowledge. Early uncertainty management systems dealt with single real values within a predefined range, soon interval valued approaches were proposed and more recently we have witnessed the introduction of fuzzy-interval valued approaches, i.e., possibility distributions and fuzzy truth-values. In this paper we describe these fuzzy set based approaches with an emphasis on the concept of fuzzy truth-value.

A new expert system (ES) to aid the nonspecialist physician in diagnosing arthritis and collagen diseases has been developed. Here we present the structure of RENOIR and the results of its implementation. This rule-based ES has been programmed using the MILORD environment. This is a shell to develop ES using a closed set of linguistic labels to express uncertainty. A feature of RENOIR is its five levels of knowledge representation, which permits to build a very flexible knowledge base (KB) and express knowledge with high accuracy. Those rules directed to similar goals are grouped in modules to improve computational performance and for higher clarity of the KB. Control of the reasoning process is assured by several mechanisms, one of the main being metarules specifically designed for almost all the knowledge levels of the KB. We have used public domain knowledge (books, criteria tables) and personal heuristics from one of the authors (Belmonte-Serrano) to implement the KB of RENOIR. In its present form, our KB comprises 1 058 rules, 978 facts, 220 metarules, and 34 modules. A first validation process has shown good performance of the ES compared to 12 physicians with diverse levels of experience in rheumatic diseases. New ongoing versions of the system with improved interfaces and reasoning capabilities are expected before verifying RENOIR's clinical acceptability. © 1994 John Wiley & Sons, Inc.

. In this paper we describe an approach to the problem of dealing with uncertainty by means of finite multi-valued logics in modular expert systems, and the results obtained. The modularity of the systems allows us to address two main characteristics of human problem-solving: the adaptation of general knowledge to particular problems and the dependency of the management of uncertainty on the different subtasks being implemented in the modules of the system, i.e. different modules can have different local multiple-valued logics as part of their local deductive mechanisms. Although the results obtained are general, we use, throughout the paper, examples of a medical expert system that has been designed using a modular language called MILORD-II, that implements them showing the practical interest of the theoretical concepts involved.

The objective of this article is to describe the MILORD Shell and particularly its architecture and its management of uncertainty. MILORD is an expert systems building tool consisting of two inference engines and an explanation module. the system allows one to perform different calculi of uncertainty on an expert defined set of linguistic terms expressing uncertainty. Each calculus corresponds to specific conjunction, disjunction, and implication operators. the internal representation of each linguistic uncertainty value is a fuzzy subset of the interval [0,1]. the different calculi of uncertainty applied to the set of linguistic terms give, as a result, a fuzzy subset that is approximated, by means of a linguistic approximation process, to a linguistic certainty value belonging to the set of linguistic terms. This linguistic approximation keeps the calculus of uncertainty closed. This has the advantage that, once the linguistic certainty values have been defined, the system computes, off-line, the conjunction, disjunction, and implication operations for all the pairs of linguistic uncertainty values in the term set and stores the results in matrices. Therefore, when MILORD is run, the propagation and combination of uncertainty is performed by simply accessing these precomputed matrices. MILORD also deals with nonmonotonic reasoning in the same framework of uncertainty management. Finally, an application to the diagnosis and treatment of pneumoniae is presented.

The issue of validity in clustering is considered and a definition of fuzzy r-cluster that extends E. Ruspini's definition (1982) is proposed. This definition is based on an indistinguishability relation based on the concept of t-norm. The fuzzy r-cluster's metrical properties are studied through the dual concept of t-conorm that leads to G-pseudometrics. From the concept of G-pseudometric, fuzzy r-clusters and fuzzy cluster coverages are defined. The authors propose a measure of cluster validity based on the concept of fuzzy coverage. The basic idea of the approach presented is that the smaller the difference between the degrees of membership and the degrees of indistinguishability, the better the clustering.>

The construction of data bases containing information concerning scientific libraries requires cooperation between experts of different scientific domains. These experts are the source of knowledge that help the librarian in the process of characterizing the documents by means of relevant features, relating them using a thesaurus and generating useful and informative classifications of the documents to facilitate its retrieval. In this work, we describe an implementation based on a knowledge acquisition process that uses a knowledge engineering methodology: Our system helps the expert in a domain, to elicitate the concepts that he uses and to characterize the documents of the domain. The system also helps the librarian in the design of the thesaurus and the classifications.