What Is Weka Data Mining?
The most known representative of freely distributed tooling is a package Weka data mining. Weka data mining represents a set of algorithms of machine training for the decision of real Data Mining-problems. Weka is written on Java and it is started practically from all platforms.
The other group of problems is presented by the tools realizing the following decisions: tools for search of associative rules, agents, valuation, regresses and forecasting, analysis of communications, consecutive patterns and temporary numbers, tools Business Intelligence, OLAP software and Database, as well as tools of transformation and data clearing. Other of them are libraries, components and tool sets for developers of creation of built in Data Mining appendices, Web Mining, which is an analysis of behaviour of sites, XML mining, search on Web, Audio and Video Mining. It is possible to say, that the popularity of some products increases, and of some of them - falls. It concerns both commercial, and freely distributed tools.
As to free-of-charge tools, in 2003, in comparison with 2002, a part of voices from tool Weka send away to the Prudsys Xelopes and R tools. In 2005 the quantity of voices for tool Weka has increased, and for Xelopes has voted essentially less users. The similar example can be given also for the commercial software: popularity of Microsoft Server SQL for Data Mining in 2003, in comparison with 2002, has increased, and in 2005 - has decreased.
The reception of reports by means of operative systems is not an optimum way. These systems spend significant time for aggregation of the information (calculation of total, average, minimal and maximal values). Besides in current base of operative system there are only the most necessary and fresh data while the information for the last periods is located in archive.
If it is necessary to receive data from archive, the duration of the report construction increases in two-three times. It is necessary to consider also, that the server of an operative system frequently does not provide necessary productivity at simultaneous construction of complex reports and input of the information.
It can catastrophically affect the work of the enterprise, as the operators cannot make out waybills, to fix shipment or reception of production when construction of the next report is carried out. The warehouse of data allows solving these problems. First, the work of a server of data warehouse does not stir to operators work. Secondly, the warehouse of data besides the detailed information contains in advance calculated aggregated values. Thirdly, in data warehouse the archival information is always accessible to inclusion in reports. All this allows to reduce considerably time of reports creation and to avoid the problems in operative work.
Terms And Conditions >> Privacy >> Contact Us
|