SwingNN

Your data is imported into a grid and used to train a neural network. The input values are forced to swing beyond their limits. The output values are forecasted by the neural network.

Pros: Trend spotting requires that one is able to discern the trends from past/historical data. Most forecasting applications that use advanced methods create and train just one neural network. SwingNN introduces a new approach to forecasting using a series of related neural networks. Multiple neural networks are created and trained. Each new neural network learns from the results produced by the previous one. The neural network that produces the most accurate results is selected and used to extrapolate and forecast results beyond the existing training range. To make the forecasts accurate the application uses multiple levels of neural networks to achieve that. Your data is imported into a grid. The grid is then used to build and train a neural network. The neural network input values are then forced to swing beyond their limits. The unknown output values are forecasted by the neural network. Then a new neural network is created using the new inputs and forecasted outputs. The two neural networks are compared. The inputs are adjusted and another new neural network is created. The process continues until a neural network produces the best results when validated. The new neural network is then used to produce more forecasts. The forecasts produced are validated and then added to the grid. All the values in the grid can be exported for you to use in any way you want.

Cons: One needs to validate the data to be absolutely sure the trend spotting is accurate enough. The other aspect is you need to be able to understand how this thing is supposed to work.

Overall: Nice set functionalities and easy to use interface. This can easily rate a 4 star.

Your data is imported into a grid and used to train a neural network. The input values are forced to swing beyond their limits. The output values are forecasted by the neural network. A new neural network is created using the new inputs and forecasted outputs. The neural networks are compared. The inputs are adjusted and another new neural network is created. The process continues until a new neural network agrees with the original about the forecasts. The forecasts are added to the grid for you to use.