In this thesis new modular neural network (MNN) architecture is proposed. The basic buildingrnblocks of the architecture are small multilayer feed forward networks trained using the Backrnpropagation algorithm (BPA). The newly proposed MNN Architecture is called Pyramidal MNNrn(PMNN). It is called Pyramidal for the number of the modules that constitutes the layers ofrnnetwork relatively decreases from the input layer to the output layer.rnAn Optimization technique called PSO has been used to optimize the topology of the proposedrnPMNN architecture for typical high dimensional input vector datasets. The optimizationrntechnique is used to suit the PMNN architecture for specific problems of high dimensional inputrnvectors depending on the nature of the data input and the nature of the problem. This is done byrnevolving topology of the modules that constitutes the network and changing the architecture ofrnthe overall network to suit the new data set.rnThe suggested training algorithm works in multiple stages depending on the number of hiddenrnlayers of the network. The training of modules in the same layer of the PMNN is easy tornimplement in parallel. Since the network is not fully connected, the number of weight ofrnconnections is less and hence the training is very quick for large input dimensional vectors.rnAn object-oriented implementation of the proposed PMNN architecture is written to simulate thernbehavior. The evaluation and optimization of the PMNN architecture for different real worldrnapplications is carried out to show the effectiveness of the proposed architecture for highrndimensional input vector applications. The evaluation is based on three pattern recognitionrnproblems: palm-print recognition, iris recognition and face recognition. In all the threernevaluations, it has achieved more than 95% accuracy of the test results. Furthermore, thernproposed PMNN architecture performs better than other similar type research works. It is shownrnthat as PMNN is a huge family of several specific architectures, this proposed topology of thernneural net can serve wide range of complex domain problems that need to be solved usingrnArtificial Intelligence (AI).rnKeywords: Modular Neural Networks, Pyramidal Modular Neural Networks, Particle SwarmrnOptimization, High Dimensional Input Vectors, General Pyramidal Modular Neural Networks,