- Fixes a bug for display of generated RBF and Sigmoidal model equations. This was a display bug only. - Maintenance.

- macOS 12 Monterey support - Runs natively on Apple Silicon and Intel - Maintenance, bug fixes, and performance improvement

Fixed a bug that prevented data tables to draw properly when scrolled macOS 11 Big Sur support

• macOS 11 support • Maintenance, bug fixes, and performance improvement

- Improved method for detecting and handling instabilities of models in ex ante system forecasting - Method for defining sets of endogenous or exogenous variables added - Maintenance, bug fixes, and performance improvement

- macOS 10.14 Mojave including dark mode appearance support - Maintenance and performance improvement

• Introducing modeling of static and dynamic, linear and non-linear systems of equations as a superset of vector auto-regressive (VAR) models for easy ex ante forecasting and self-detection of the interdependence structure of the modeled system - System models: Self-organized regression equations of all models of a system are added to the model report - System model graph: The relevance of a model input is now displayed when selected • New examples added: - System models of the U.S., Japanese, and German national economies with ex ante forecasts until 2021, compared to forecasts provided by the International Monetary Fund Example - Forecasting of financial market indexes - System model for an air quality forecasting problem • Maintenance and performance improvement

• Introducing modeling of static and dynamic, linear and non-linear systems of equations as a superset of vector auto-regressive (VAR) models for easy ex ante forecasting and self-detection of the interdependence structure of the modeled system - System models: Self-organized regression equations of all models of a system are added to the model report - System model graph: The relevance of a model input is now displayed when selected • New examples added: - System models of the U.S., Japanese, and German national economies with ex ante forecasts until 2021, compared to forecasts provided by the International Monetary Fund Example - Forecasting of financial market indexes - System model for an air quality forecasting problem • Maintenance and performance improvement

- Improved print handling for plots, reports, and tables. - Fixes a bug that may appear when exporting dynamic models to Excel. - General maintenance.

- Adds new information to cost curve diagrams. - Fixes two bugs in reporting classification and cost-optimized models.

- Fixes a bug with Microsoft Excel for Mac, version 15, when trying to import data from a worksheet. - General maintenance and performance update.

- Fixes a synchronisation bug when adding new data to worktables.

- Fixes a bug with Microsoft Excel for Mac, version 15, when trying to import data from a worksheet. - General maintenance and performance update.

macOS 10.12 Sierra support

*** Promotional prices until May 6, 2016 *** - New model complexity optimization option to get most parsimonious, most compact models in result of automated model self-organization from data. - Generation of an explicit model equation from data of the form y = f(x) for linear, multi-linear, and non-linear models, which represents new extracted knowledge and which significantly helps interpreting models. - For multi-linear models, Insights now ensures that the obtained final model is full multi-linear, not only the neuron transfer functions. - Calculation of the importance of the model inputs for both model composites and individual models developed with Insights version 5. - New plots added for composite and individual models developed with Insights version 5: - Importance of model inputs and variables for all parametric models; - ROC and Cost Curves as valuable tools for binary classification models. - Maintenance and performance optimization.

- Significantly improved processing speed by new In-Memory database; option between fast In-Memory DB or file-based DB for large amount of data records. - New model export to Python. Integrate ready-to-use Python code of your developed models into your web applications or other projects. - Display of model accuracy of predictions in model plots on-the-fly as an immediate measure of model performance on new data. - For classification models, additional display of key discrimination parameters for easy evaluation of the power of the model on predicted data. - Updated and new examples added: - Gene expression of 2000 genes (inputs) for modeling tumor tissue as an example for underdetermined, high-dimensional modelling; - Network intrusion detection as an example for the problem of detecting rare events; - Ionosphere modification and analysis of pulsed high-power radar signals from a phased array of antennas; - Daily ozone level detection. - General maintenance and bug fixing.

- Maintenance and bug fixing for OS X 10.11 El Capitan. - A new modeling option that allows you to replace the current model with the the new model has been added. This reduces the size of the modelbase and makes it more manageable. - An additional parameter, the Area under the ROC Curve, often used to rank classification models is now calculated and reported for all classification models. - The default number of training and forecast data to work with, which is copied to a model after modeling, can be set now manually in the preference pane. - Labeling of plots has been redesigned.

- Maintenance and stability release. - New feature for self-detection respectively manual specification of stationarity of time processes for Similar Patterns. - Improved self-optimization of Similar Patterns composites.

- Feature for self-detection respectively manual specification of stationarity of time processes for Similar Patterns added. - Improved self-optimization of Similar Patterns composites. - Option for deleting all models in a model base added. - New example on Congressional voting behavior added. - Maintenance and stability release.

Mac OS X 10.10 update.

- Fixed model export bugs on very complex models exported to Excel (static and classification models) and AppleScript (RBF models). - Maintenance.

- New example on Wind Farm Energy Generation added. - Maintenance.

• New example on wind farm power generation added. • Fixed a bug in version 3.1.2 that in some cases avoided the generation of higher accurate models.

• In version 3 Similar Patterns has been implemented as an outstanding modeling technology that finds self-similar patterns in complex, fuzzy, or very noisy time processes for very fast and easy forecasting such as market prices, market demand, or sales figures. • Version 3.1 adds the Autoupdate function for Similar Patterns models. Run simulations on historical data stored in Excel for backtesting, for example, or use it on real-time data for instant forecasting. All results are reported and stored in Excel for further application. • Adds two new model export formats for integrating models into workflows or projects directly: In addition to implementing models ready-to-use in Excel, now export your models working in AppleScript or in TEXT format for further analysis in MATLAB, for example. • New example models and data added; • Maintenance and performance improvements;

OS X 10.9 support.