|
StatConsulting developed AdvancedMiner as an advanced data analysis tool. It supports the complete range of tools required for the building of Data Mining models and advanced data analyses. AdvancedMiner is the currently developed version of software package known formerly as Górnik System.
The application provides tools for: importing, processing and transformation of data, statistical analysis and its visualization, creation of various analytical models (approximation, classification, segmentation, association and survival models), model quality assessment, application of the models to data, exporting of models in the form of scoring code, reporting.
For constructing models a selection of algorithms, available as separate modules, can be used. These include Classification Trees, Linear Regression, Logistic Regression, IRLS, Bivariate Probit, Neuron Networks, Kohonen Networks SOM-s, K-Means, A-priori Algorithms, Discriminant Analysis, Time Series, Probability Distribution Library, Optimization Library, Matrix Calculus Library, and Monte Carlo Methods Library. The main assets of the AdvancedMiner applicationA wide array of analytical tools provided by the application: Provides the means to construct various analytical models, including ones supporting Churn, Cross/Up selling, customer segmentation, LTV, and Survival analyses.
Integrated model building environment: Allows for easy comparisons of different model creation methods. Supports easy construction of hybrid models, combining a number of different analytical methods. Streamlines the adjustment of existing modules to the needs of the customer and to the design of new specialized solutions. Guarantees conformity with Data Mining and data processing standards, which facilitates integration with other systems.
Distributed computing: Accelerates the analysis process while fully utilizing the available computational resources. Decreases the cost of building an efficient and scalable computer system. Accommodates the use of low cost PC class machines running different operating systems.
Productive user interface: Models can be designed with the help of wizards or with the use of scripts. The scripting environment supports multifaceted data processing, including SQL programming. Multiple databases can be connected and available at the same time. Reports can be generated in all popular formats: MS Office, OpenOffice, HTML, and PDF. A uniform object-oriented approach allows for intuitive application of various methods, which decreases the time required for learning about new features. Data and models can be visualized in a transparent and interactive manner.
If you are interested in licensing or learning more about our software, please feel free to contact us. |