Inference in expert systems based on complete multivalued logic
Influence of modeling structure in probabilistic sequential decision problems
Markov Decision Processes (MDPs) are a classical framework for stochastic sequential decision problems, based on an enumerated state space representation. More compact and structured representations have been proposed: factorization techniques use state variables representations, while decomposition techniques are based on a partition of the state space into sub-regions and take advantage of the resulting structure of the state transition graph. We use a family of probabilistic exploration-like...
Informal proofs formally checked by machine
Information boundedness principle in fuzzy inference process
The information boundedness principle requires that the knowledge obtained as a result of an inference process should not have more information than that contained in the consequent of the rule. From this point of view relevancy transformation operators as a generalization of implications are investigated.
Innovative applications of associative morphological memories for image processing and pattern recognition.
Morphological Associative Memories have been proposed for some image denoising applications. They can be applied to other less restricted domains, like image retrieval and hyperspectral image unsupervised segmentation. In this paper we present these applications. In both cases the key idea is that Autoassociative Morphological Memories selective sensitivity to erosive and dilative noise can be applied to detect the morphological independence between patterns. Linear unmixing based on the sets of...
Instability of the eikonal equation and shape from shading
Instability of the eikonal equation and shape from shading
In the shape from shading problem of computer vision one attempts to recover the three-dimensional shape of an object or landscape from the shading on a single image. Under the assumptions that the surface is dusty, distant, and illuminated only from above, the problem reduces to that of solving the eikonal equation |Du|=f on a domain in . Despite various existence and uniqueness theorems for smooth solutions, we show that this problem is unstable, which is catastrophic for general numerical algorithms. ...
Integrated region-based segmentation using color components and texture features with prior shape knowledge
Segmentation is the art of partitioning an image into different regions where each one has some degree of uniformity in its feature space. A number of methods have been proposed and blind segmentation is one of them. It uses intrinsic image features, such as pixel intensity, color components and texture. However, some virtues, like poor contrast, noise and occlusion, can weaken the procedure. To overcome them, prior knowledge of the object of interest has to be incorporated in a top-down procedure...
Integration of candidate hash trees in concurrent processing of frequent itemset queries using Apriori
Integration of expert knowledge into a probabilistic expert system
Intelligent decision-making system for autonomous robots
The paper gives an account of research results concerning a project on creating a fully autonomous robotic decisionmaking system, able to interact with its environment and based on a mathematical model of human cognitive-behavioural psychology, with some key elements of personality psychology included. The principal idea of the paper is focused on the concept of needs, with a certain instrumental role of emotions.
Intelligent Nesting System
Interpolative fuzzy algorithms for intelligent controllers and decision making
Interpretability of linguistic variables: a formal account
This contribution is concerned with the interpretability of fuzzy rule-based systems. While this property is widely considered to be a crucial one in fuzzy rule-based modeling, a more detailed formal investigation of what “interpretability” actually means is not available. So far, interpretability has most often been associated with rather heuristic assumptions about shape and mutual overlapping of fuzzy membership functions. In this paper, we attempt to approach this problem from a more general...
Interpretable random forest model for identification of edge 3-uncolorable cubic graphs
Random forest is an ensemble method of machine learning that reaches a high level of accuracy in decision-making but is difficult to understand from the point of view of interpreting local or global decisions. In the article, we use this method as a means to analyze the edge 3-colorability of cubic graphs and to find the properties of the graphs that affect it most strongly. The main contributions of the presented research are four original datasets suitable for machine learning methods, a random...
Interpretation and optimization of the -means algorithm
The paper gives a new interpretation and a possible optimization of the well-known -means algorithm for searching for a locally optimal partition of the set which consists of disjoint nonempty subsets , . For this purpose, a new divided -means algorithm was constructed as a limit case of the known smoothed -means algorithm. It is shown that the algorithm constructed in this way coincides with the -means algorithm if during the iterative procedure no data points appear in the Voronoi diagram....
Interpretation of pattern classification results, obtained from a test set
The present paper presents and discusses a methodology for interpreting the results, obtained from the application of a pattern classifier to an independent test set. It addresses the problem of testing the random classification null hypothesis in the multiclass case, by introducing an exact probability technique. The discussion of this technique includes the presentation of an interval estimation technique for the probability of correct classification, which is slightly more accurate than the ones...
Interval analysis for certified numerical solution of problems in robotics
Interval analysis is a relatively new mathematical tool that allows one to deal with problems that may have to be solved numerically with a computer. Examples of such problems are system solving and global optimization, but numerous other problems may be addressed as well. This approach has the following general advantages: (a) it allows to find solutions of a problem only within some finite domain which make sense as soon as the unknowns in the problem are physical parameters; (b) numerical computer...
Interval-valued linguistic summaries of databases
Intrinsic dimensionality and small sample properties of classifiers
Small learning-set properties of the Euclidean distance, the Parzen window, the minimum empirical error and the nonlinear single layer perceptron classifiers depend on an “intrinsic dimensionality” of the data, however the Fisher linear discriminant function is sensitive to all dimensions. There is no unique definition of the “intrinsic dimensionality”. The dimensionality of the subspace where the data points are situated is not a sufficient definition of the “intrinsic dimensionality”. An exact...