TiModel Predict is designed to predict the future market trends.
It is developed mainly for intraday trading (М1, М5 and М15), but it works on all timeframes.
The indicator does not redraw history and value on a zero bar. It allows you to use the product for developing various strategies.
Due to normalizing methods employed, the indicator is capable of working on all markets and instruments.
How it works
Quotes are used to form models with different dynamic characteristics. They are displayed in the chart as broken lines. Dynamics of the models can be modified by varying the threshold value in the settings. Several trend lines are plotted to increase the efficiency of trading decisions. Since the market is constantly changing, the performance of some models might get worse or better as the time goes by. So a trader should choose the most suitable model for a given situation and perform an analysis.
It is based on the operation of a formalized neural network. The network analyzes several previous break points and gives a prediction of a future trend direction (future ray). The indicator generates "TiModel_Predict.csv" file containing values of the last and the next break point.
- LengthSample, bars - the length of historical data in bars.
- Threshold, pips - the pattern dynamics (threshold) in pips.
- Recommended Threshold, pips - recommended threshold in pips. This parameter is dynamic. It varies and depends on the symbol and the history length.
- The horizontal line does not carry any information. It is drawn for the sake of convenience to keep the last break point in sight.
- The optimal length for intraday trading is 500-1500 bars.
- The threshold value has an impact on the number of break points. The smaller the threshold, the more often the indicator will draw lines, and vice versa.
- The indicator demonstrates good results in a moderate steady market. When there are strong market movements caused by important news releases, it is advisable to use additional signals (analysis of senior timeframes, multicurrency) when taking trade decisions.
Fixed names of input parameters