System modeling is the task of constructing representative models of processes and has become an invaluable tool in many different areas of science and engineering. Due to the inherent complexity of many real world systems the application of traditional techniques is limited. In such instances more sophisticated (so called intelligent) modeling approaches are required. Fuzzy Neural modeling is one such technique, which by integrating the attributes of fuzzy systems and neural networks is ideally suited to system modeling. Unfortunately, the application of these systems is limited to low dimensional problems for which good quality expert knowledge and data are available. The work described in this thesis reflects new learning algorithm based on Takagi-Sugeno fuzzy model structure, three basic steps are used to tune the parameters in the neurofuzzy model, the means of the membership functions tuned using C-Mean Clustering algorithm, the variances and the rule weights were tuned using non-linear least square method (the membership type used in this work is gauss membership function). The developed construction algorithm successfully identifies parsimonious models, as a result algorithm, which is more efficient in time but decreasing the accuracy of the identified model. The method introduced in this thesis is illustrated on many different examples, including multi-dimensional static and dynamical systems. The modeling results are compared with those of the MATLAB package (ANFIS daptive-TVetwork-based Fuzzy Inference System). It's found that those of the proposed algorithm are better in forms of computational time and accuracy.