Optimization for Distributed Committee Machines in The Knowledge Discovery in Distributed Databases Process

The knowledge discovery in distributed databases represents the overall process of obtaining useful information from data that is stored in a replication topology on multiple computing systems. Distributed committee machines are formed by many neural networks that work in a distributed manner on multiple computing systems for resolving one data mining task. These architectures are the most suitable neural structures for working in a replication topology. In my case I used multilayer perceptrons trained with the backpropagation algorithm for resolving the classification task. In this paper I propose an optimized way of functioning for the entire distributed committee machine architecture by choosing the appropriate database engine, the best type of replication and by reducing the writing operations into the distributed database as much as possible. All the experiments were done on large data sets in order to achieve the best results. This study is useful in research areas such as distributed learning, adaptive eLearning, machine learning, medical diagnosis, astronomy and many others.