Generalization bounds for the regression of real-valued functions

This paper suggests a new bound of estimating the confidence interval defined by the absolute value of difference between the true (or general) and empirical risks for the regression of real-valued functions. The theoretical bounds of confidence intervals can be derived in the sense of probably approximately correct (PAC) learning. However, these theoretical bounds are too overestimated and not well fitted to the empirical data. In this sense, a new bound of the confidence interval which can explain the behavior of learning machines more faithfully to the given samples, is suggested.