Navegación

Búsqueda

Búsqueda avanzada

Tuning Neural Networks in a Fuzzy Logic Programming Environment

Wide datasets are usually used for training and validating neural networks, which can be later tuned in order to correct their behaviors according to a few number of test cases proposed by users. In this paper we show how the FLOPER system developed in our research group is able to perform this last task after coding a neural network with a fuzzy logic language where program rules extend the classical notion of clause by including on their bodies both fuzzy connectives (useful for modeling activation functions of neurons) and truth degrees (associated to weights and bias in neural networks). We present an online tool which helps to select such operators and values in an automatic way, accomplishing with our recent technique for tuning this kind of fuzzy programs. Moreover, we provide some experimental results revealing that our tool generates the choices that better fit user’s preferences in a very efficient way, and producing relevant improvements on tuned neural networks.

Tuning Fuzzy Logic Programs with Symbolic Execution (Trabajo de alto nivel)

Fuzzy logic programming is a growing declarative paradigm aiming to integrate fuzzy logic into logic programming. One of the most difficult tasks when specifying a fuzzy logic program is determining the right weights for each rule, as well as the most appropriate fuzzy connectives and operators. In this paper, we introduce a symbolic extension of fuzzy logic programs in which some of these parameters can be left unknown, so that the user can easily see the impact of their possible values. Furthermore, given a number of test cases, the most appropriate values for these parameters can be automatically computed.