The input signals of brain computer interfaces may be either electroencephalogram recorded from scalp or electrocorticogram recorded with subdural electrodes. It is very important that the classifiers have the ability for discriminating signals which are recorded in different sessions to make brain computer interfaces practical in use. This paper proposes a method for classifying motor imagery electrocorticogram signals recorded in different sessions. Extracted feature vectors based on wavelet transform were classified by using k-nearest neighbor, support vector machine and linear discriminant analysis algorithms. The proposed method was successfully applied to Data Set I of BCI competition 2005, and achieved a classification accuracy of 94 % on the test data. The performance of the proposed method was confirmed in terms of sensitivity, specificity and Kappa and compared with that of other studies used the same data set. This paper is an extended version of our work that won the Best Paper Award at the 33rd International Conference on Telecommunications and Signal Processing.