Resumen
The article presents an overview of knowledge discovery or the computer techniques that, in the broadest sense, automatically find fundamental properties and principles. Dealing with more comprehensive, non-oversimplified, real-world types of data, such as mixed data types, is a key direction for the future. It emphasizes the aspects of knowledge discovery that are shared across the various fields of science. Although computers can't crank out major scientific findings every day, they promise to accelerate the rate of significant scientific discoveries. Selected Techniques covers four especially important techniques for knowledge discovery. The notion of computers automatically finding useful information is an exciting and promising aspect of just about any application intended to be of practical use. Replacing such creative human activities as sorting through it all with a computer represents a grand human endeavor. Based on a simulation of the architecture and functioning of the human brain, neural networks are being applied to many types of problems, including pattern recognition, process control, and real-world prediction of real-world phenomena