This paper presents a new architecture of neural networks designed for pattern recognition. The concept of induction graphs coupled with a divide-and-conquer strategy defines a Graph of Neural Network (GNN). It is based on a set of several little neural networks, each one discriminating only two classes. The principles used to perform the decision of classification are: a branch quality index and a selection by elimination. A significant gain in the global classification rate can be obtained by using a GNN. This is illustrated by tests on databases from the UCI machine learning database repository. The experimental results show that a GNN can achieve an improved performance in classification.