|
The three Alyuda neural networks-based tools are targeted toward users who have different goals. All of them are designed to solve real-world problems. All share similar GUI oncepts and Alyuda proprietary heuristics. To find out which of the products suits you best, use the feature comparison table below.
|
| Feature |
Alyuda Forecaster XL |
Alyuda Forecaster |
Alyuda NeuroIntelligence |
| General |
| Excel add-in interface (optimized for MS Excel users) |
 |
|
|
| Wizard-like interface (different modes for beginners and experts) |
|
|
|
| Windows tabbed interface (optimized for experts) |
|
|
|
| Automatic and manual data analysis and preprocessing |
|
|
|
| Automatic selection of neural network architecture and training parameters |
|
|
|
| Online help system |
|
|
|
| Free technical support |
|
|
|
| Sample financial, business and scientific problems included |
|
|
|
| Analyze and Pre-process Your Data |
| Import popular ASCII file formats (CSV, TXT, PRN) |
|
|
|
| Import Excel files) |
|
|
|
| Custom date formats and file structure definition |
|
|
|
| Automatic data analysis and pre-processing |
|
|
|
| Automatic categorical values encoding |
|
|
|
| Automatic numeric values scaling |
|
|
|
| Automatic Date/Time values encoding |
|
|
|
| Manual min/max values specification for scaling |
|
|
|
| Visual representation of data anomalies |
|
|
|
| Outliers handling for numeric data (customizable outlier coefficient) |
|
|
|
| Missing values handling for numeric values (removal and 4 substitution options) |
|
|
|
| Missing values handling for categorical values (removal and 3 substitution options) |
|
|
|
| Automatic recognition of data entry errors (wrong type values) |
|
|
|
| Detailed data analysis and data preprocessing reports |
|
|
|
| Automatic dataset partition to training, validation and test sets (random or sequential) |
|
|
|
| Manual dataset partition to training, validation and test sets |
|
|
|
| Manual column type identification (numeric, categorical, date, time, text) |
|
|
|
| Accept/ignore records and columns manually |
|
|
|
| Preprocessed data representation |
|
|
|
| Binary columns for anomalies indication |
|
|
|
| Two methods of automatic lag columns insertion |
|
|
|
| Statistical information for data columns |
|
|
|
| Design Neural Network |
| Input feature selection (GA, stepwise, exhaustive). |
|
|
|
| Fully automated neural network design with a constructive algorithm. |
|
|
|
| Fully automated neural network design using architecture search heuristics |
|
|
|
| Manual architecture specification (for multi-layer perceptron) |
|
|
|
| Customizable heuristic architecture search method |
|
|
|
| Three heuristic methods of neural network architecture search. |
|
|
|
| Exhaustive architecture search with customizable parameters |
|
|
|
| Customizable search range and search sensitivity |
|
|
|
| Detailed statistics for each tested architecture |
|
|
|
| Network fitness criteria: AIC, Test set error, Correlation, R-squared |
|
|
|
| Graphical representation of network fitness |
|
|
|
| Time-series networks |
|
|
|
| Network visualization |
|
|
|
| Network sets |
|
|
|
| Automatic adjustment of learning rate and momentum for Back-Propagation algorithm |
|
|
|
| Training algorithms: Conjugate Gradient Descent, Levenberg-Marquardt, Quick-Propagation, Incremental and Batch Back-Propagation |
|
|
|
| Additional training algorithms: Quasi-Newton, Quasi-Newton (Limited Memory) |
|
|
|
| Activation functions: Linear, Logistic, Tanh, Softmax |
|
|
|
| Error functions: Sum-of-Squares, Cross-entropy |
|
|
|
| Classification model: Winner-takes-all, Confidence-limits (Accept/Reject levels) |
|
|
|
| Heuristics for automatic generation of stop training conditions |
|
|
|
| Generalization loss control (10 preset levels) |
|
|
|
| Retrain network to get better results |
|
|
|
| Manual stopping conditions (target error level, error improvement, correct classification rate, number of iterations) |
|
|
|
| Real-time control on training parameters (MSE, MAE, CCR, # of iterations). |
|
|
|
| Training Error Graph (network error by iteration) |
|
|
|
| Training Error Table (network error and error improvement by iteration) |
|
|
|
| Control Network Training Process |
| Real-time output of training parameters |
|
|
|
| Continue training with new parameters |
|
|
|
| Jog weights |
|
|
|
| Add jitter |
|
|
|
| Correlation and r-squared real-time graphs |
|
|
|
| Error improvement graph |
|
|
|
| Weights distribution graph |
|
|
|
| Error distribution graph |
|
|
|
| Input importance graph |
|
|
|
| Training log: test and validation set error for each iteration |
|
|
|
| Early-stopping on generalization loss |
|
|
|
| Retain and restore best network |
|
|
|
| Automatic network retrains and selection of the best network among retrains |
|
|
|
| Manual network retrain |
|
|
|
| Retrains statistics |
|
|
|
| Weights initialization: manual randomization range; optimized for Uniform or Gaussian distribution |
|
|
|
| Test and Analyze Performance |
| Actual vs Forecasted graph |
|
|
|
| Actual vs Forecasted scatter plot |
|
|
|
| Confusion matrix |
|
|
|
| Response graph |
|
|
|
| ROC curve |
|
|
|
| Actual vs Forecasted table with absolute and relative errors |
|
|
|
| Tolerance levels to quickly estimate overall forecasting quality |
|
|
|
| Input importance graph |
|
|
|
| Estimated forecasting error |
|
|
|
| Apply Network |
| Enter new cases manually or from the Clipboard |
|
|
|
| Load new cases from a new data file |
|
|
|
| Apply to selected records from your original dataset |
|
|
|
| Visual output representation with Response Graph |
|
|
|
| Output representation with Results Table |
|
|
|
| Confidence limits for network output |
|
|
|
| Save results in a separate file |
|
|
|
| Enjoy User Interface Extras |
| Detailed explanations on every step |
|
|
|
| Customizable reports (with preview and printing capabilities) |
|
|
|
| Reports export to HTML and XLS |
|
|
|
| Save/Load neural network |
|
|
|
| Two convenient methods of data selection in one interface (by range and by column) |
|
|
|
| Complete color customization for reports and graphs |
|
|
|
| Neural network autosave |
|
|
|
| Price |
| Single User Edition |
$ 149 |
$ 179 |
$ 399 |
| Unlimited Site Edition |
$ 999 |
|
$ 3999 |
