Getting a forecast, Steps to make a forecast, Neural Nets (NN) basics, NN Outputs, Detrended Zigzag, NN Inputs, Training the Neural Net, Stop and the training process, Auto stop, Move the NN projection line to the Main Window
Now, after learning the basics of the software (which is ability to download data, manipulate objects on the screen and work with main modules of the program), we are ready to discuss its core idea. Timing Solution is focused on getting a market forecast. As a result, we should have a projection line, with evaluation of its reliability.
The process of getting a reliable projection line consists of two parts:
a) creating a model that serves as a base for the forecast;
b) making the forecast.
With Timing Solution software, it is easy to create a model, due to the software's unique feature, Universal Language of Events (you can read about it here).
Let say, we have already found a model that describes previous market conditions rather well. It is time to make a forecast (to create a projection line) based on that model. Timing Solution software does it by applying Neural Network technology. You do not need to be a professional in Neural Nets. It is enough to remember several things mentioned below.
Click the mouse here:
You enter the part of the program where you will be able to create a forecast based on any event/events. Whatever model you apply, the forecast is made by Neural Net. It is a special math technology that studies how your chosen model has performed in the past and uses that knowledge to tell what is more likely to happen in the future.
Let us discuss the major items of Neural Net forecasting while creating a projection line based on Ptolemy aspects of transiting planets. This procedure will take about 5 minutes of your time. Follow these steps:
1) Run Neural Network module
2) Click here (to define the outputs - things to be forecasted):
As an example, let us forecast the detrended oscillator with the period of 50 bars:
It is called RPO50 - relative price oscillator with the period of 50 bars. Detrending is necessary to do, as Neural Network works better with detrended indicators (it looks for real connections between the price movement of your financial instrument and things that form your model; the existence of some trend confuses this search).
3) Click here (to define a model used as a base of the forecast):
Thus we define the Ptolemy aspects that are used to forecast RPO50 (our inputs). Let us do it with the orb of 15 degrees.
4) Now it is the time for Neural Network to start a learning procedure. A learning procedure is the way to train our Neural Net. We do that on the training interval (which is located before LBC). Click here:
5) In seconds you will get something like this:
As you see, the projection line (a red curve) describes the RPO50 (a black line) very good inside the blue (training) interval. That was our goal - to train the Neural Net till it finds the parameters of our model that make it fit well to the price oscillator.
You may want to see right now how this model is performing after the training interval. You can do that choosing any portion of the price chart after the time interval used for training. In order to do that, highlight the time interval by dragging the mouse cursor over it (press the left mouse button in the beginning of this interval -> move the mouse cursor to the end of the interval keeping left mouse button pressed -> release left mouse button):
To get the idea how our projection line fits the price (to be exact, the price oscillator), look at this panel with the calculated correlation coefficient between the price and the projection line (Neural Network and Linear projection lines):
It shows the correlation calculated for the selected interval. While you choose different intervals, the program recalculates this information panel, so you can see how the projection line fits the price on different intervals.
6) When you decide that Neural Work projection line is OK (it fits the price quite well on different intervals), click "Stop" button to break the training procedure:
You may set auto stop feature here:
In this example Neural Network stops training when the correlation between the price and the projection line created by Neural Network reaches 15% at least.
You can define more complicated auto stop criteria:
Here Neural Network stops training when the correlation between Neural Network projection line and the price reaches 15% at least OR when the correlation between Linear projection line and the price reaches 36% at least.
Another example involves the amount of training steps:
Here the program stops Neural Network training when the correlation reaches 15% OR the amount of training steps reaches 100K steps.
Remember that typing 15% means that we indicate the correlation, while typing 100K means the amount of training steps.
7) Follow these steps to put Neural Network's projection line into the Main window and hide the results panel:
(We do hide the Results panel because it covers the Main window. Do not worry, you are able to see this panel any time).
8) Here it is, our projection line (i.e. forecast) in the Main window: