We elaborate on hierarchical credal sets, which are sets of probability mass functions paired with second-order distributions. A new criterion to make decisions based on these models is proposed. This is achieved by sampling from the set of mass functions and considering the Kullback-Leibler divergence from the weighted center of mass of the set. We evaluate this criterion in a simple classification scenario: the results show performance improvements when compared to a credal classifier where the second-order distribution is not taken into account.

Recently, representations and methods aimed at analysing decision problems where probabilities and values (utilities) are associated with distributions over them (second-order representations) have been suggested. In this paper we present an approach to how imprecise information can be modelled by means of second-order distributions and how a risk evaluation process can be elaborated by integrating procedures for numerically imprecise probabilities and utilities. We discuss some shortcomings of the use of the principle of maximising the expected utility and of utility theory in general, and offer remedies by the introduction of supplementary decision rules based on a concept of risk constraints taking advantage of second-order distributions.

The sensory perceptions from humans are intertwined channels,which assemble diverse data in order to decrypt emotionalinformation. Just by associations, humans can mix emotionalinformation, i.e. emotion detection through facial expressionscriteria, emotional speech, and the challenging field of emotionalbody language over the body poses and motion. In this work, wepresent an approach that can predict six basic universal emotionscollected by responses linked to human body poses, from acomputational perspective. The emotional outputs could be fedas inputs to a synthetic socially skilled agent capable ofinteraction, in the context of socially intelligent systems. Themethodology uses a classification technique of information fromsix images extracted from a video, entirely developed using themotion sensing input device of Xbox 360 by Microsoft. We aretaking into account that the emotional body language containsadvantageous information about the emotional state of humans,especially when bodily reaction brings about consciousemotional experiences. The body parts are windows that showemotions and they would be particularly suitable to decodingaffective states. The group of extracted images is merged in oneimage with all the relevant information. The recovered image willserve as input to the classifiers. The analysis of images fromhuman body poses makes it possible to obtain relevantinformation through the combination of proper data in the sameimage. It is shown by experimental results that the SVM candetect emotion with good accuracy compared to other classifiers.

This paper analyzes the integration of neural networks and linear systems for the identification, state estimation and output feedback control of weakly nonlinear systems. Considering previous knowledge about the system given by approximated linear state-space models, linear observers and linear controllers, training algorithms for the neuro-identification, state neuro-estimation and output feedback neuro-control were derived considering the dynamics of the nonlinear system. It was found that the integrated linear-neuro model can identify the dynamics of the system much more accurately than a purely linear model or a purely neuro model. It was also found that the state estimation and vibration isolation performance of the system with integrated linear-neuro output feedback control is better than the system with linear control or neuro-control.

The point of view of Isaac Asimov is unlikely in a close future, but machines that develop tasks in a sensible manner are already a fact. In light of this remark, recent research tries to understand the requirements and design options that imply providing an autonomous agent with means for detecting emotions. If we think about of exporting this model to machines, it is possible that they become capable to evolve emotionally according to such models and would take part in the society more or less cooperatively, according to the perceived emotional state. The main purpose of this research is the implementation of a decision model affected by emotional feedback in a cognitive robotic assistant that can capture information about the world around it. The robot will use multi-modal communication to assist the societal participation of persons deprived of conventional modes of communication. The aim is a machine that can predict what the user will do next and be ready to give the best possible assistance, taking in account the emotional factor. The results indicate the benefits and importance of emotional feedback in the closed loop human-robot interaction framework. Cognitive agents are shown to be capable of adapting to emotional information from humans.

The ultimate objective of studying, modeling and analyzing policy problems is to incorporate the newest management technologies in the public policy decision-making in a meaningful and practically feasible way that adds significant value to the process. Simulation techniques can support the policy decision process by allowing empirical evaluation of the system dynamics present in the policy situation at hand. This paper presents a decision support simulation model for the European Union (EU) Climate and Energy targets 2030 as a case study of public policy decision making on the EU level. The simulation model is based on the problem structuring or framing by derivation of a system dynamics model from verbal descriptions of the problem, the graphical representation and analysis of change scenarios using the ‘Causal Mapping and Situation Formulation’ method. This approach supports the analysis of qualitative and quantitative information in order to facilitate both the conceptualization and formulation stages of the system modeling process. The resulting model, which is simply a topology of quantified causal dependencies among the problem key variables, can be used to simulate the transfer of change. The aim of simulation herein is to apply cognitive strategic thinking and scenario-based planning in a public policy problem situation in order to design alternative options and provide foresight or ex-ante impact assessment in terms of economic, social, environmental and other impacts.

Strategic planning models and information provision for decision-making in complex strategic situations are frequent subjects for scientific research. This research deals with the problem of supporting strategic planning decision-making in public higher education (HE) institutions by designing a Decision Support System (DSS) to be used by HE decision makers in implementing their strategic planning process, considering that the DSS would be anchored in on all databases of the institution’s information systems. This paper adopts a model for the strategic planning process, advocates the incorporation of technologies of participation (ToP) and introduces a collaborative framework for the planning activities at the different institutional levels to develop the institution’s strategic plan using a bottom-up approach. Based on the strategic planning process model, a DSS framework is proposed and decision support methods are suggested for the different modules of the DSS. The DSS provides intelligent support (on the individual, group and organizational levels) to strategic planning decisions in all stages of the process. By utilizing this DSS, it is possible to create better conditions for implementing the objectives of the future-oriented activity of the institution.

Multi-objective optimization is a way to manage multiple objectives in analytical decision support systems. However, for real-life problems, different types of uncertainty often become prominent when defining the model. In this paper, we analyze these different types of uncertainties and suggest a suitable typology for a decision process based upon multi-objective optimization models. Uncertainty analysis can be performed based on the proposed typology; therefore, this analysis provides the necessary support for a decision maker in the identification the crucial uncertainty in the decision process.

Traditional multi-objective optimization attempts to find Pareto optimal solutions. Since a Pareto optimal set can be huge, the problem of selecting one or few solutions occurs. Post optimality analysis in multi-objective optimization requires incorporation of decision makers' preferences in the form of weights. In this paper the concept of robustness with regards to weights is introduced. The different types of weights robustness show how sensitive a solution is to variation in weights' coefficients. An approach for analysis of Pareto optimal sets through weights robustness is then devised. The suggested approach can be of special interest in the presence of conflicting preferences among decision makers or when preference information is unavailable. In conclusion, managerial usage it in the different strategies for negotiation provides possibility to thoroughly weigh all alternatives before settling on an agreement.

We utilize second-order probability distributions for modeling second-order information over imprecise evidence in the form of credal sets. We generalize the Dirichlet distribution to a shifted version, denoted the S-Dirichlet, which allows one to restrict the support of the distribution by lower bounds. Based on the S-Dirichlet distribution, we present a simple combination schema denoted as second-order credal combination (SOCC), which takes second-order probability into account. The combination schema is based on a set of particles, sampled from the operands, and a set of weights that are obtained through the S-Dirichlet distribution. We show by examples that the second-order probability distribution over the imprecise joint evidence can be remarkably concentrated and hence that the credal combination operator can significantly overestimate the imprecision.

This paper analyzes the performance and practical implementation of fuzzy-neural networks for the autonomous motion of mobile robots. The designed fuzzy-neural controller is a refined version of a conventional fuzzy controller, and was trained to optimize a given cost function minimizing positioning error. It was found that the mobile robot with fuzzy-neural controller presents good positioning and tracking performance for different types of desired trajectories. It was verified by computer simulation as well as experimentally using a laboratory-scale car-like robot model.

Positioning and tracking control systems are an important component of autonomous robot applications. This paper presents the design method of tracking control systems based on H infinite preview control where the present and future desired positions of the robot are used to determine the control actions to be applied so that the robot describes the desired trajectory as close as possible. The performance improvements achieved with H infinite preview control have been examined in the frequency and time domains for different types of reference signals when applied to a one-dimensional positioning system. It was found that preview control improves the tracking performance by improving the phase response of the tracking system.

For robots to plan their actions autonomously and interact with people, recognizing human emotions is crucial. For most humans nonverbal cues such as pitch, loudness, spectrum, speech rate are efficient carriers of emotions. The features of the sound of a spoken voice probably contains crucial information on the emotional state of the speaker, within this framework, a machine might use such properties of sound to recognize emotions. This work evaluated six different kinds of classifiers to predict six basic universal emotions from non-verbal features of human speech. The classification techniques used information from six audio files extracted from the eNTERFACE05 audio-visual emotion database. The information gain from a decision tree was also used in order to choose the most significant speech features, from a set of acoustic features commonly extracted in emotion analysis. The classifiers were evaluated with the proposed features and the features selected by the decision tree. With this feature selection could be observed that each one of compared classifiers increased the global accuracy and the recall. The best performance was obtained with Support Vector Machine and bayesNet.

This paper focuses on a system of recognizing human’s emotion from a detected human’s face. The analyzed information is conveyed by the regions of the eye and the mouth into a merged new image in various facial expressions pertaining to six universal basic facial emotions. The output information obtained could be fed as an input to a machine capable to interact with social skills, in the context of building socially intelligent systems. The methodology uses a classification technique of information into a new fused image which is composed of two blocks integrated by the area of the eyes and mouth, very sensitive areas to changes human’s expression and that are particularly relevant for the decoding of emotional expressions. Finally we use the merged image as an input to a feed-forward neural network trained by back-propagation. Such analysis of merged images makes it possible, obtain relevant information through the combination of proper data in the same image and reduce the training set time while preserved classification rate. It is shown by experimental results that the proposed algorithm can detect emotion with good accuracy.

Adequate representation of imprecise probabilities is a crucial and non-trivial problem in decision analysis. Second-order probability distributions is the model for imprecise probabilities whose merits are discussed in this thesis.

That imprecise probabilities may be represented by second-order probability distributions is well known but there has been little attention to specific distributions. Since different probability distributions have different properties, the study of the desired properties of models of imprecise probabilities with respect to second-order models require analysis of particular second-order distributions.

An often held objection to second-order probabilities is the apparent arbitrariness in the choice of distribution. We find some evidence that the structure of second-order distributions is an important factor that prohibits arbitrary choice of distributions. In particular, the properties of two second-order distributions are investigated; the uniform joint distribution and a variant of the Dirichlet distribution that has the property of being the normalised product of its own marginal distributions.

The joint uniform distribution is in this thesis shown to have marginal distributions that belie the supposed non-informativeness of a uniform distribution. On the other hand, the modified Dirichlet distribution discovered here has its information content evenly divided among the joint and marginal distributions in that the total correlation of the variables is minimal.

It is also argued in the thesis that discrete distributions, as opposed to the continuous distributions mentioned above, would have the advantage of providing a natural setting for updating of lower bounds, and computation of expected utility is made more efficient.

We present a notion, relative independence, that models independence in relation to a predicate.  The intuition is to capture the notion of a minimum of dependencies among variables with respect to the predicate. We prove that relative independence coincides with conditional independence only in a trivial case. For use in second-order probability, we let the predicate express first-order probability, i.e. that the probability variables must sum to one in order to restrict dependency to the necessary relation between probabilities of exhaustive and mutually exclusive events. We then show examples of Dirichlet distributions that do and do not have the property of relative independence.  These distributions are compared with respect to the impact of further dependencies, apart from those imposed by the predicate.

Since second-order probability distributions assign probabilities to probabilities there is uncertainty on two levels. Although different types of uncertainty have been distinguished before and corresponding measures suggested, the distinction made here between first- and second-order levels of uncertainty has not been considered before. In this paper previously existing measures are considered from the perspective of first- and second-order uncertainty and new measures are introduced. We conclude that the concepts of uncertainty and informativeness needs to be qualified if used in a second-order probability context and suggest that from a certain point of view information can not be minimized, just shifted from one level to another.

The success of a development intervention is primarily defined by its longer-term results. Information and Communication Technology for Development (ICT4D) projects are no exception. In particular, the external context in which such projects are to be deployed plays a crucial role for achieving better development results. Thus, an analysis of contextual factors is essential for decision-makers at the project proposal screening and evaluation. Guided by the results-based management framework, a tool for influence assessment of contextual factors is designed and presented in this paper. The tool comprises a suggested model for common ICT4D contextual factors and a complementing method for their assessment. The method employs a scenario- based judgments elicitation process for data extraction together with Monte- Carlo simulation for data processing and aggregation. While the tool is not intended for comprehensive and deep analysis, the underlying streamlined process along with the easy to interpret assessment results arguably ensure further wide adoption.

The long-term success or failure of a development project is largely shaped by the external context. Therefore, assessment of factors influencing fulfilment of long-term development outcomes is vital for better project planning. In recent decades, the logical framework (logframe) has de facto become a standard tool for planning and managing development interventions. While the logframe requires identification of assumptions and risks regarding the external context, it does not suggest ways to analyse them in a conventional risk assessment manner. Also, the log-frame has been criticised for ignoring uncertainty in project environment along with neglecting external opportunities. Therefore, in this paper we suggest a method for project context analysis that extends the log-frame with scenarios analysis and address aforementioned shortcomings. We implement and demonstrate the application of the method on an international aid development project, discuss the method's potential use-cases, specific limitations and future development.

Strategic fit is a crucial criterion for screening strategically important projects. However, this aspect is vague and unquantifiable to employ with conventional quantitative analysis. This study suggests a method to assess strategic fit as relevance of a project to a set of organisation strategic goals in the context of multi-criteria project evaluation. The assessment is done subjectively by means of linguistic values, which are translated into numerical ones by employing a fuzzy logic.