• Overview
  • Reviews
  • Comments

SSA Stochastic

Stochastic oscillator with Singular Spectrum Analysis algorithms (SSA).

SSA is an effective method of treatment of non-stationary time series with unknown internal structure. It is used for the determination of the main components (trend, seasonal fluctuations and wave), smoothing and noise suppression. The method allows us to find previously unknown series periodicities and make forecasts on the basis of the detected periodic patterns.

The indicator signals are identical to the original signal indicator but have an important advantage - they do not have a time delay with respect to the price dynamics and more accurately and synchronously reflect the variability of the behavior of the price series. This is achieved by the fact that the "SSA-% D" signal is realized not by moving average, but using the low-pass filtering algorithm using the SSA. Thus, this indicator does not lag behind.

Adjustable noise filtering can significantly reduce the number of false alarms that are typical of the original indicator.

The forecast for "SSA-% K" and "SSA-% D" takes into account the totality of the detected different scale factors that determine the behavior of the price series and can be used to reduce risks in the strategy.

The characteristic behavior of signals and interpretation of the indicator correspond to the same properties of the Stochastic oscillator.

For the convenience of the user, both the original "% K" and "% D" and upgraded "SSA-% K" and "SSA-% D" calculations are provided here.

 

Options

  1. K period period of observations
  2. D period smoothing period
  3. Slowing repeated smoothing period
  4. Warning ranges ranges of warning signals
  5. Algorithm prediction method
  6. N: Data fragment fragment number of "%K" for analysis
  7. Time-dependent lag window effect of history on the value at the point
  8. % K high-frequency limit noise filtering option for the treatment of "%K".
  9. % D frequency limit setting smoothing and filtering for the construction of "%D".
  10. Forecast smoothing smoothing/regularization forecast
  11. Recalculate period period of updating the indicator values
  12. Predictable Points number of points of the prediction.
  13. BackwardShift shift back a history fragment for setting up the model and the forecast according to the known data.
  14. INTERFACE/Magic Number ID for the application of the indicator within an EA
  15. VISUAL OPTIONS "SSA-% K" and "SSA-% D" graph color settings.


Explanation of the choice of parameters

High frequence limit determines the level of filtering and suppression of random noise in the data. Oscillations whose contributions do not exceed this level will be filtered out.

BacwardShift for setting the indicator for a particular set of data. Specifying the offset allows us to compare the forecast with known values and select the indicator settings more accurately.

Forecast smoothing balance the results of the projection, suppressing "outliers" or by using "weighting factors", taking into account the importance of previous results.

NOTE: If you choose "weight" option of smoothing the forecast, the four assessments of the forecast values are performed on the first indicator calculation step to initialize the smoothing stack.

Therefore, the first step requires more computing time.  It is not required in the next steps.

Magic Number. Connecting the results of the indicator to an EA is possible when requesting 8 ("% K" -Original), 9 ("% D" -Original), 10 ("SSA-% K") , 11 ("SSA-% D")  and 12 ("Warning Signal") buffers.

No reviews