Session Program


  • 11 July 2017
  • 04:00PM - 06:00PM
  • Room: Giardino
  • Chairs: Timothy C. Havens, Christian Wagner and Derek T. Anderson

Innovations in Fuzzy Inference

Abstract - As non-singleton fuzzy logic controllers (NSFLCs) are capable of capturing input uncertainties, they have been effectively used to control and navigate unmanned aerial vehicles (UAVs) recently. To further enhance the capability to handle the input uncertainty for the UAV applications, a novel NSFLC with the recently introduced similarity-based inference engine, i.e., Sim-NSFLC, is developed. In this paper, a comparative study in a 3D trajectory tracking application has been carried out using the aforementioned Sim-NSFLC and the NSFLCs with the standard as well as centroid composition-based inference engines, i.e., Sta-NSFLC and Cen-NSFLC. All the NSFLCs are developed within the robot operating system (ROS) using the C++ programming language. Extensive ROS Gazebo simulation-based experiments show that the Sim-NSFLCs can achieve better control performance for the UAVs in comparison with the Sta-NSFLCs and Cen-NSFLCs under different input noise levels.
Abstract - The state-of-art algorithms in computational intelligence have become better than human intelligence in some of pattern recognition areas. Most of these state-of-art algorithms have been developed from the concept of multi- layered artificial neural networks. Large amount of numerical and linguistic rule data has been created in recent years. Fuzzy sets are useful in modeling uncertainty due to vagueness, ambiguity and imprecision. Fuzzy inference systems incorporate linguistic rules intelligible to human beings. Many attempts have been made to combine assets of fuzzy sets, fuzzy inference systems and artificial neural networks. Use of a single fuzzy inference system limits the performance. In this paper, we propose a generic architecture of multi-layered network developed from Takagi Sugeno fuzzy inference systems as basic units. This generic architecture is called {$\backslash$}enquote\{Takagi Sugeno Deep Fuzzy Network\}. Multiple distinct fuzzy inference structures can be identified using proposed architecture. A general three layered TS deep fuzzy network is explained in detail in this paper. The generic algorithm for identification of all network parameters of three layered deep fuzzy network using error backpropagation is presented in the paper. The proposed architecture as well as its identification procedure are validated using two experimental case studies. The performance of proposed architecture is evaluated in normal, imprecise and vague situations and it is compared with performance of artificial neural network with same architecture. The results illustrate that the proposed architecture eclipses over three layered feedforward artificial neural network in all situations.
Abstract - Over the last few years, both the study and the design of IT implementations of CAM have gained a renewed interest. The success of these models in the Theory of Modelling and Simulation (TMS) relies on the structural phenomenon of emergence which makes it possible to run realistic simulations, despite lacking a modeling process for real systems. CAMs do not describe real systems with complex equations, they allow the complexity of real systems to emerge from simple interactions described locally from their cellular elements. In order to optimize simulations whatever the spatial dimension considered, the concept of activity is used. In this work, we introduce disturbances in propagation rules and we improve simulation rendering. We express a doubt in the expression of the cell's activity, i.e. we express the activity rule by means of an Fuzzy Inference System (FIS). We present a new way to use FIS, in an activity-based cellular modeling approach for fire spreading simulations.
Abstract - The fuzzy integral (FI) is a nonlinear aggregation operator whose behavior is defined by the fuzzy measure (FM). As an aggregation operator, the FI is commonly used for evidence fusion where it combines sources of information based on the worth of each subset of sources. One drawback to FI-based methods, however, is the specification of the FM. Defining the FM manually quickly becomes too tedious since the number of FM terms scales as 2\^{}n, where n is the number of sources; thus, an automatic method of defining the FM is necessary. In this paper, we review a data-driven method of learning the FM via minimizing the sum-of-squared error (SSE) in the context of decision-level fusion and propose an extension allowing knowledge of the underlying FM to be encoded in the algorithm. The algorithm is applied to real-world and toy datasets and results show that the extension can improve classification accuracy. Furthermore, we introduce a visualization strategy to simultaneously show the quantitative information in the FM as well as the FI.
Abstract - In classical logic, Modus Ponens allows to infer new knowledge in the case where the antecedent of a given rule is observed, establishing that the rule conclusion then holds. Approximate reasoning extends the principle to the case where the observation does not totally match the rule antecedent. Several approaches have been proposed to deal with the extreme case where the observation is actually disjoint from the rule antecedent, using different principles to guide inference and avoid producing total uncertainty. This paper studies two of them, namely Geometric Compatibility Modification (GCM) and the Transformation-based Constraint-Guided Generalised Modus Ponens (T-CGMP), that respectively perform a type of approximate analogical reasoning and extend the GMP: it provides an indepth comparison, to determine their relationships, common points and distinct features. It thus provides guidelines for the definition of fuzzy inference schemes.
Abstract - Numerous applications in engineering are plagued by incomplete data. The subject explored in this article is how to extend the fuzzy integral (FI), a parametric nonlinear aggregation function, to missing data. We show there is no universally correct solution. Depending on context, different types of uncertainty are present and assumptions are applicable. Two major approaches exist, use just observed data or model/impute missing data. Three extensions are put forth with respect to just use observed data and a two step process, modeling/imputation and FI extension, is proposed for using missing data. In addition, an algorithm is proposed for learning the FI relative to missing data. The impact of using and not using modeled/imputed data relative to different aggregation operators-selections of underlying fuzzy measure (capacity)-are also discussed. Last, a case study and data-driven learning experiment are provided to demonstrate the behavior and range of the proposed concepts.