Evaluation of a New Incremental Classification Tree Algorithm for Mining High Speed Data Streams

Almost all the current computer applications are so designed to produce the huge volume of diverse data. Analysis and reorganization of such data require a huge computer memory than available. At present, new machine learning algorithms are being designed and developed to overcome these problems. Even though the incremental learning algorithms are effective in saving time and memory, they cannot be used to process larger datasets. Recently, the researchers are concentrating on the development of new algorithms, which are suitable for processing such large datasets. The algorithms are so designed that they enable the computer to learn from a single pass, require less memory usage and update incrementally at a later stage. Further, they should perform the data mining task at any time.