http://www.gmdhshell.com/docs/ 2012-05-18T03:13:22-04:00 http://www.gmdhshell.com/docs/ http://www.gmdhshell.com/docs/lib/images/favicon.ico text/html 2011-11-05T14:05:28-04:00 http://www.gmdhshell.com/docs/airline_passengers?rev=1320516328&do=diff Airline passengers This is an example of time series forecasting. Original source of data is a book by Box and Jenkins (1976): Times Series Analysis: Forecasting and Control, p. 531. The dataset consists of 144 monthly totals of airline passengers from January 1949 to December 1960. text/html 2011-11-04T11:37:11-04:00 http://www.gmdhshell.com/docs/batch_mode_export?rev=1320421031&do=diff Batch mode export Running GMDH Shell from the command line turns on the batch-mode export. There is only one plug-in available for the batch-mode export called Export Forecast. No additional command line keys are required for batch-mode exporting. text/html 2012-05-15T19:08:24-04:00 http://www.gmdhshell.com/docs/changelog?rev=1337123304&do=diff Changelog Version 2.6 (2012-05-16) * Renewed GUI: Toolbars, Data Manager tab and Model browser. * Save charts as images (bitmaps - PNG and vectors - EMF), zoom out, switch to full-screen - now within the context menu (right-click) for all charts. * Date/time variable now can be analyzed using country-specific Holiday calendars. * Support for datasets with future data. (For example you have weather forecast or a schedule for data points you wish to forecast. Thus, these columns will be … text/html 2012-02-08T08:08:44-04:00 http://www.gmdhshell.com/docs/command_line?rev=1328706524&do=diff Command line Background processing In the background mode GMDH Shell performs modeling simulation without any interaction. The easiest way to launch the program in the background mode is to use the command line key /auto Use of the key /auto executes the project located in the current directory. So, the usual way for background processing is to configure a project visually and then run it from the command line. Example: ”“GMDH Shell.exe” /auto” No other settings required. text/html 2012-03-30T05:35:45-04:00 http://www.gmdhshell.com/docs/concepts?rev=1333100145&do=diff Concepts Here you'll find a list of concepts and definitions important for understanding and successful application of GMDH Shell. Graphical user interface (GUI) GMDH Shell is a Windows application, its GUI consists of a host application and plug-ins that draw their panels inside the host window. There are two areas - tabs and the sidebar where plug-in panels can be placed. text/html 2011-08-27T10:06:57-04:00 http://www.gmdhshell.com/docs/control_panel?rev=1314454017&do=diff . Control panel Control panel gives you quick access to the most frequent operations. It has a tool bar with buttons and a progress bar (see at the screenshot). or while the simulation is in progress Import data. Show data file in default editor. Start processing. Stop processing. Replaces the Start button while processing. Save models. Load models. Apply models (scoring). Settings. Save project settings. Wizard. Help. Registration. text/html 2011-11-10T16:12:56-04:00 http://www.gmdhshell.com/docs/export_excel_formula?rev=1320959576&do=diff How to export a formula to Excel Click the green button to start the playback: Download with the embedded player: [How_to_export_excel_formula.zip] Other videos: How to import and process data Usability hints text/html 2011-11-03T21:28:11-04:00 http://www.gmdhshell.com/docs/import?rev=1320370091&do=diff The import module reads data from source files. CSV/XLS/XLSX Import CSV/XLS/XLSX module reads data from .csv (text files with delimeters) and spreadsheets .xlsx and .xls. Limitations: The module reads only from the first sheet of .xlsx or .xls file. The module can't read from password-protected files. text/html 2011-08-27T18:53:49-04:00 http://www.gmdhshell.com/docs/iris_flower?rev=1314485629&do=diff Iris flower This is an example of multi-class classification. Data source is a classical dataset introduced by Sir Ronald Aylmer Fisher (1936) as an example of discriminant analysis (Wikipedia article). The data set consists of 150 instances and 5 variables. It was modified so that the last 30 instances are equally represented by 3 Iris flower species. text/html 2012-01-24T06:46:48-04:00 http://www.gmdhshell.com/docs/learning_algorithms?rev=1327405608&do=diff Learning algorithms Learning algorithm is a procedure that controls model optimization process. In the modern theory of predictive modeling it is well known that the model should provide a trade-off between simplicity and accuracy. GMDH Shell matches this goal using learning algorithms of the Group Method of Data Handling (GMDH) that is a knowledge discovery technique originated by Prof. A.G.Ivakhnenko in 1968. text/html 2011-11-03T21:03:52-04:00 http://www.gmdhshell.com/docs/menu?rev=1320368632&do=diff Menu The menu bar divides operations available in the host window into several categories. Most frequent operations are duplicated at the Control panel. There are 5 categories: File - Common operations with files, templates, and settings. text/html 2012-02-11T20:23:29-04:00 http://www.gmdhshell.com/docs/model_browser?rev=1329009809&do=diff Model browser Every modeling simulation produces a set of predictive models ranked by their testing accuracy, but visualization panels of GMDH Shell show characteristics of only one model at a time. The Model browser allows you to switch between these models and choose which one to be displayed in visualization panels. text/html 2012-01-05T08:07:27-04:00 http://www.gmdhshell.com/docs/modeling_with_gmdh_shell?rev=1325768847&do=diff User Guide GMDH Shell is a predictive modeling tool that allows to setup quickly and efficiently and quick setup to several important and popular data analysis and forecasting tasks allowing not only professional data-analysts to design predictive simulations and automate forecasting tasks. text/html 2011-02-06T19:46:33-04:00 http://www.gmdhshell.com/docs/one-click_examples?rev=1297039593&do=diff [Airline passengers][Top500 supercomputers] [Iris flower][Synthetic dataset] text/html 2012-02-08T15:29:40-04:00 http://www.gmdhshell.com/docs/plug-in_modules?rev=1328732980&do=diff Plug-in modules Plug-in modules are building blocks of the data processing chain. There are 6 types of modules: Picture Module type Available modules [Import] Import 1 [Preprocess] Preprocess 2 [Solver] Solver 1 [Postprocess] Postprocess 1 [Visualization] Visualization 1 [Export] Export 1 text/html 2011-11-04T11:04:49-04:00 http://www.gmdhshell.com/docs/postprocess?rev=1320419089&do=diff [Postprocess panel] This type of plug-ins is currently represented by a single module. Average of top-ranked models Averaging of predictions from a number of top-ranked models usually reduces variance of residuals. Averaging of models sometimes decreases accuracy of predictions. text/html 2012-02-09T19:33:45-04:00 http://www.gmdhshell.com/docs/preprocess?rev=1328834025&do=diff Preprocess Preprocess plug-in transforms data according to requirements of the Solver module. Data manager The Data manager is a visual tool that allows a user to specify columns of data-files which will serve as model inputs and prediction targets. text/html 2011-10-10T21:03:02-04:00 http://www.gmdhshell.com/docs/sidebar?rev=1318294982&do=diff Back to gmdhshell.com Table of contents Getting started * Video tutorials * Learning algorithms * Concepts * Templates * One-click examples Host application * Menu * Control panel * Model browser * Command line Plug-in modules text/html 2011-12-15T07:18:54-04:00 http://www.gmdhshell.com/docs/solver?rev=1323951534&do=diff Solver is a plug-in module that solves a modeling problem and returns a set of predictive models. You can accept solver settings loaded form a template or configure it manually. In the solver configuration panel you can do the following * Optionally, expand input variables with automatically-generated artificial variables - Additional variables. * Optionally, shuffle data rows in order to make statistical characteristics of training and testing data more uniformly - Dataset reordering. * … text/html 2011-12-30T10:19:10-04:00 http://www.gmdhshell.com/docs/start?rev=1325258350&do=diff GMDH Shell is a predictive modeling software that provides tools for data exploration, modeling, forecasting, measuring out-of-sample accuracy and application of selected models to new data. Notable features of GMDH Shell Solving modeling problems: * Multivariate time series forecasting * Regression (continuous value prediction) * Classification (prediction of a category) * Ranking and selection of variables * Polynomial curve fitting text/html 2011-08-27T18:55:29-04:00 http://www.gmdhshell.com/docs/synthetic_dataset?rev=1314485729&do=diff Synthetic dataset This is an example of regression problem. Three variables were randomly generated in a spreadsheet and coupled with output of polynomial function y = 100*x1 + x1*x2 – x2^2 that uses only two variables. All columns of the obtained dataset were noised by up to 10% (uniformly distributed random error between -10 and 10%). text/html 2012-04-04T15:04:17-04:00 http://www.gmdhshell.com/docs/templates?rev=1333566257&do=diff Time series forecasting template The general template for time series forecasting is appropriate for multivariate and univariate time series analysis and forecasting. First of all you should select Input and Target variables in your data and if necessary transform them. text/html 2011-08-27T18:54:58-04:00 http://www.gmdhshell.com/docs/top500_supercomputers?rev=1314485698&do=diff Top500 supercomputers This is an example of time series forecasting. Data source is The Top500 List of Supercomputer Sites released semiannually by www.top500.org. Every list contains information about different parameters of modern supercomputers including rank of a computer system, theoretical and measured computational power, the number of processing cores, manufacturer information, component types and models, operating system, etc. Example data is a historical list of entry level supercomp… text/html 2011-11-03T15:24:31-04:00 http://www.gmdhshell.com/docs/tutorial01_video?rev=1320348271&do=diff How to import and process data Click the green button to start the playback: Download with the embedded player: [How_to_import_and_process_data.zip] Other videos: How to export Excel formula Usability hints text/html 2011-04-25T14:37:19-04:00 http://www.gmdhshell.com/docs/tutorials?rev=1303756639&do=diff Video tutorials How to import and process data How to export Excel formula Usability hints text/html 2011-11-04T00:27:00-04:00 http://www.gmdhshell.com/docs/usability_hints?rev=1320380820&do=diff Usability hints Summary: * You can move panels from the sidebar to tabs and vise-versa, or simply hide a panel. * Selected fields can be copied to clipboard. It can be a row or a column in a table, all cells, a model formula, etc. * Ctrl+A - select all * Ctrl+C - copy to clipboard * Ctrl+V - paste from clipboard text/html 2012-02-12T14:54:18-04:00 http://www.gmdhshell.com/docs/visualization?rev=1329076458&do=diff After the computation has finished, visualization helps to explore computational results using the plot, the table of predictions (Table), the table of variables' importance (Importance), the performance panel and a special tool called model browser.