Course Name: Trading with Machine Learning: Regression, Section No: 3, Unit No: 5, Unit type: Notebook
We did all the calculation on raw price as moving average and close minus open and everything else .... then we scale the data after? Feels like it makes no sense. It feels like all calculations should be done on the rescaled data. As it is shown, we get scaled versions of the original moving average.
Hi David,
The order in which calculations and scaling are performed depends on the specific context and the goals and requirements of the analysis. Typically, calculations are performed on the raw price data before scaling because scaling the data can affect the values of the calculations. Performing calculations on the raw data allows you to preserve the relationships and patterns within the data before scaling. Scaling can then be applied to put the data on a standardized scale. This makes it easier to compare different stocks or time periods.
Additionally, some calculations, such as technical indicators, are often designed to be used with unscaled data. For example, the Relative Strength Index (RSI) is calculated using the ratio of average gains to average losses, which is based on the raw price data. Scaling the data before calculating the RSI could result in inaccurate values.