In this chapter, we've discussed a few topics that make a neural network work better, either by improving its accuracy or by extending its knowledge. These techniques help a lot in designing solutions with artificial neural networks. The reader is welcome to apply this framework in any desired task that neural networks can be used on, in order to explore the enhanced power that these structures can have. Even simple details such as selecting the input data may influence the entire learning process, as well as the filtering of bad data or the elimination of redundant variables. We demonstrated in two implementations, two strategies that help to improve the performance of a neural network: stochastic online learning and adaptive resonance theory. These methodologies enable the network to extend its knowledge and therefore, adapt to new changing environments.