The present invention provides a method for entropy decoding a video signal, the method comprising the steps of: obtaining a context model initial value associated with a current slice; calculating a probability value on the basis of a syntax statistic of a previous slice; deriving weight values with respect to the context model initial value associated with the current slice and the probability value calculated on the basis of the syntax statistic of the previous slice; and updating the context model initial value associated with the current slice using the weight values.