Neural Designer is a software tool created to provide data scientists with results in a way previously unachievable.

Today, most organizations are stuck at lower-value descriptive analytics. But more sophisticated predictive and prescriptive analytics can bring greater business value. This is known as advanced analytics.

In recent years, neural networks are dramatically improving the data analytics field. These are artificial intelligence models which have the ability to automatically lean from data sets.

However, data scientists have trouble applying advanced neural networks. The main reasons for that are the high complexity of algorithms and their difficulty of use.

Neural Designer is a software tool that has been developed to meet that market needs: it implements advanced neural network algorithms and is very easy to use.

It contains a graphical user interface that clearly defines the workflow and provides comprehensive results.

In summary, Neural Designer allows you to get actionable insights resulting in smarter decisions and better business outcomes.

Our typical customers include analytics departments at innovative companies, big data consulting firms and research centers.

Innovative companies

Consulting firms

Research centers

With Neural Designer there is no need for programming or building complicated block diagrams. Its user interface guides the user through a sequence of well-defined steps, in order to simplify data entry.

The software comes with many tutorials and examples to help you learn how to use it, and our team provides specialized technical support to our customers through telephone, email and Skype.

The software creates dashboards displaying comprehensive results from all tasks. Users can explore data and visualize the results with tables, charts and pictures that can be exported and used in other tools.

Also, the whole report can be exported to PDF format.

The objective is that the user is able to understand and interpret each step in the analytics process.

Neural Designer contains a large range of advanced neural network algorithms that allow data scientists to build the most powerful predictive models.

Sophisticated methods for data pre-processing, such as cleaning outliers or calculating principal components, are included.

The software implements neural network architectures with an arbitrary number of nonlinear layers, in order to build the most powerful predictive models.

It also includes different error and regularization methods so that the user can achieve the best results for a given data set.

Sophisticated training algorithms, such as the quasi-Newton method or the Levenberg-Marquardt algorithm, have been developed for more accurate computations.

One of the main advantages of Neural Designer is the inclusion of an advanced model selection framework, which allows the user to obtain the most relevant variables in a given process.

The tool also contains many different methods for testing exhaustively the generalization capabilities of a predictive model, depending on the application at hand.

Finally, the resulting neural network can be operated within the program in different ways, or exported to any programming language, such as R or Python, using its mathematical expression.

Neural Designer outstands in terms of performance. It is developed in C++ for better memory management and higher processing speed, and implements CPU parallelization by means of OpenMP and GPU acceleration with CUDA.

The following plots show the results of a performance comparison among a predictive analytics software in R, a data mining platform in Java and Neural Designer.

As we can see, Neural Designer is able to analyze data sets up to 1000 times bigger, is up to 9 times faster and is up to 100 times more accurate its main competitors.

Neural Designer provides an easy way for deploying predictive models. For that, you can use a standard such as PMML, or export the resulting model to programming languages such as R or Python.

An example of that expression in the R programming language is written below.

expression <- function(x) { scaled_x<-2*(x+1)/(1+1)-1 y_1_1<-tanh(0.361707-0.497807*scaled_x) y_1_2<-tanh(-0.15776-0.376231*scaled_x) y_1_3<-tanh(0.295148+0.493422*scaled_x) scaled_variable_2<-(-0.0531044-0.868102*y_1_1-0.778027*y_1_2+0.818376*y_1_3) outputs <- c(0.5*(scaled_variable_2+1.0)*(1+1)-1) outputs }

- Approximation (or modelling) to discover intricate relationships.
- Classification (or pattern recognition) to recognize complex patterns.

- Compatible with the most common data files: Excel, OpenOffice, CSV, Weka, DAT, TXT …
- Also compatible with the most common data bases: Oracle, MySQL, SQLite, SQL Server and Access.
- Complete configuration of variables and instances.
- Exhaustive descriptive statistics.
- Estimation of variables importance by means of linear/logistic correlations.
- Advanced methods for data balancing.
- Innovative utilities for outlier detection.

- Network architecture with unlimited number of layers.
- Threshold, symmetric threshold, logistic, hyperbolic tangent and linear activation functions.
- Scaling and unscaling layers with minimum/maximum and mean/standard deviation methods.
- Probabilistic layer with binary and softmax methods.

- Sum squared error, mean squared error, root mean squared error and normalized squared error functionals for common data sets.
- Minkowski error for dealing with outliers.
- Cross-entropy error for pattern recognition.
- Weighted squared error for unbalanced data sets.
- Regularization for avoiding overfitting.
- Gradient descent and conjugate gradient for training of big data sets.
- Quasi-Newton method for fast training of medium data sets.
- Levenberg-Marquardt algorithm for very fast training of small data sets.

- Incremental order and simulated annealing for finding the optimal network architecture.
- Growing inputs, pruning inputs and genetic algorithm for selecting the most important features.

- Complete error data and corresponding statistics calculation.
- Linear regression analysis for function regression problems.
- Confusion matrix for pattern recognition applications.
- Full set of metrics for evaluation of binary classifiers.
- ROC curve for diagnostic tests.
- Cumulative gain and lift charts for segmentation applications in marketing.
- Calibration plot for classification problems.
- Error autocorrelation and cross-correlation for time series prediction.
- List of misclassified instances.

- Calculation of output values.
- Directional plots for exploring the predictive model.
- Jacobian values.
- Exportable mathematical expression of the model.
- Exportable predictive model in R and Python.

- Exhaustive results in an interactive report plenty of descriptions, tables and figures.
- Report exportable to Word and Pdf formats.
- Results exportable in data format.

- Extensive tutorials on the application of neural networks with Neural Designer.
- Several examples of machine learning applications in different fields.
- Premium technical support by email, phone or video chat.

- Software developed with the high-performance programing language C++.
- Code subjected to optimization techniques for memory management and processing speed.
- CPU parallelization by means of OpenMP.
- GPU acceleration with CUDA.

- Windows.
- Mac OS X.
- Linux (Debian and Ubuntu).

- Amazon Web Services.
- Microsoft Azure.