With the immense horizontal and vertical growth of the World Wide Web (WWW), it is becoming more popular for website owners to showcase their innovations, business, and concepts. Along side they are also interested in tracking and understanding the need of the users. Analyzing web access logs, one can understand the browsing behavior of users. However, web access logs are voluminous as well as complex. Therefore, a semi-automatic intelligent analyzer can be used to find out the browsing patterns of a user. Moreover, the pattern which is revealed from this deluge of web access logs must be interesting, useful, and understandable. A radial basis function neural networks (RBFNs) with mixture of kernels are used in this work for classification of web access logs. In this connection two RBFNs with different mixture of kernels are investigated on web access logs for classification. The collected data are used for training, validation, and testing of the models. The performances of these models are compared with RBFNs. It is concluded that mixture of appropriate kernels are an attractive alternative to RBFNs.