Neural Network and Fuzzy Time Series. Forecasting using neural network and fuzzy time series

      Аннотация: This work deals with neural networks (NN), specifically with multi-layered NN from the algorithm learning point of view. We will describe feed forward neural network (FFNN), recurrent neural network (RCNN) and introduce basic facts about NN, which will be used later in dissertation. A neural network is a mathematical model that is inspired by biological neural networks and tries to simulate them. It consists of interconnected units - neurons, which are the computation units of a neural network. NNs are part of Artificial Intelligence. The knowledge is stored in connections between neurons which are called synaptic weights (weights), simplification of biological dendrites and axons. NN is a universal aproximator of relations stored inside of data - a nonlinear statistical data modeling aproximator, is able to learn and adapt its structure based on internal/external information that is propagated through NN during learning phase. It is relatively easy to use in wide area of technical and nontechnical areas without further theoretical knowledge for most of NNs. There is a number of NNs that require knowledge to implement them and use correct set of initialization parameter.

Ключевые слова: ANN, DSS, Fuzzy time series, Fuzzy Set, Forecasting, high order model, Production Linguistic Value