Feedforward nets with sigmoidal activation functions are often designed by minimizing a cost criterion. It has been pointed out before that this technique may be outperformed by the classical perceptron learning rule, at least on some problems. In this paper, we show that no such pathologies can arise if the error criterion is of a threshold LMS type, i.e., is zero for values “beyond” the desir...

We consider the existence of efficient algorithms for learning the class of half-spaces in ~n in the agnostic learning model (Le., making no prior assumptions on the example-generating distribution). The resulting combinatorial problem finding the best agreement half-space over an input sample is NP hard to approximate to within some constant factor. We suggest a way to circumvent this theoreti...

Digital straight lines can be charactarized by Rosenfeld's chord property as well as Hung's evenness property. Both properties can be formulated in terms of linear inequalities. Because linear inequalities can be veriied by the predicates of a Gamba perceptron, this leads in a straightforward manner to the design of perceptrons that can recognize digital straight lines. In this paper we will di...

Using the cavity method we consider the learning of noisy teachergenerated examples by a nonlinear student perceptron. For insufficient examples and weak weight decay, the activation distribution of the training examples exhibits a gap for the more difficult examples. This illustrates that the outliers are sacrificed for the overall performance. Simulation shows that the picture of the smooth e...

In this paper, a comparison study is presented on artificial intelligence and time series models in 1-hour-ahead wind speed forecasting. Three types of typical neural networks, namely adaptive linear element, multilayer perceptrons, and radial basis function, and ARMA time series model are investigated. The wind speed data used are the hourly mean wind speed data collected at Binalood site in I...