User:Melquiades81/Rulex

Rulex is a software platform developed by the company Rulex, Inc., that provides an integrated environment for data exploration, machine learning, data mining, predictive analytics and business analytics. It is used for business and industrial applications as well as for research, education, training, rapid prototyping, and application development and supports all steps of the data mining process including results visualization, validation, and optimization.

History
The idea behind the Rulex system dates back to 1991, when Marco Muselli, researcher at the Italian National Research Council began work on efficient methods for Boolean function synthesis. The first algorithm of the Rulex suite was completed in 1998, but the first commercial applications were exploited in 2003, when an e-nose system based on Rulex was designed. In 2007, Impara Srl, a spin-off of the Italian National Research Council, was founded to trade the software. In the meanwhile, an intuitive Graphical User Interface was added to enlarge to user target of the methods. Since 2013, Rulex is sold by the American company Rulex Inc.

Description
Rulex uses a client–server model with the server offered as ...

Rulex main interface is based on a graphical block diagramming tool that enables access to several tasks of data processing and analysis. Each block, performing a specific analysis, belongs to one of the following categories:


 * Visualization-editing: blocks for viewing and elaborating the data and the results of the analysis. This group include the Data Manager, a task for modifying the data, performing queries (filters, groups, sorts) by simple drag-and-drops, drawing plots and computing statistics.
 * Preprocessing: methods for properly preparing the data before the out-and-out analysis. This group include: ETL (Extract Transform Load) modules, time series management and data reshaping.
 * Classification: methods to induce the classification of labeled samples. This group include Logic Learning Machine (LLM), a proprietary algorithm for inducing rules from a dataset.
 * Clustering: methods for determining the partition in groups of unlabeled samples.
 * Regression: methods for obtaining the relation between several ordered variables.
 * Association Rules: methods for finding patterns frequently associated with the same identifier.
 * Optimization: methods for optimizing the classification results according to the desired constraints.
 * Evaluation: methods for evaluating the results of the analysis

Blocks can be connected by wires to perform the desired sequence of analysis. Each wire carries all the amount of data (stream) regarding the studied problem, so that every block has always the needed information at its disposal. Constraints on the possible connection between blocks are imposed to avoid unconsistency in the stream of data. The data (and the results) at each stage of the analysis are stored in a database in a compressed format to reduce the memory occupation. In this way, the flow of information can quickly be reconstructed at any time. Currently, Rulex supports these database platforms: SQLServer, MySql, SQLite, PostgreSQL, Oracle.