Skip to main content
Cross Validated

Return to Question

improved formatting
Source Link
Michael R. Chernick
Michael R. Chernick
  • 43.2k
  • 28
  • 85
  • 159

How toshould I normalize a time series data that is highly volatileshows high volatility in order to be used in an ANN? Let's take the price of Bitcoin asfor example.

I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison butand also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as the same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series datalikedata like the price of Bitcoin?

How to normalize a time series data that is highly volatile to be used in an ANN? Let's take the price of Bitcoin as example.

I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison but also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series datalike the price of Bitcoin?

How should I normalize time series data that shows high volatility in order to be used in an ANN? Let's take the price of Bitcoin for example.

I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison and also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as the same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series data like the price of Bitcoin?

improved formatting
Source Link
edit approved Aug 27, 2018 at 19:38
Seanny123
edit approved Aug 27, 2018 at 19:38
Seanny123
  • 667
  • 4
  • 21

The question is: How to normalize a time series data that is highly volatile to be used in an ANN? Let's take the price of Bitcoin as example.
I

I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison but also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series datalike the price of Bitcoin?

The question is: How to normalize a time series data that is highly volatile to be used in an ANN? Let's take the price of Bitcoin as example.
I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison but also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series datalike the price of Bitcoin?

How to normalize a time series data that is highly volatile to be used in an ANN? Let's take the price of Bitcoin as example.

I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison but also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series datalike the price of Bitcoin?

Tweeted twitter.com/StackStats/status/1025985048904957953
Source Link
Manngo
Manngo
  • 467
  • 4
  • 12

How to normalize highly volatile time series?

The question is: How to normalize a time series data that is highly volatile to be used in an ANN? Let's take the price of Bitcoin as example.
I have the following doubts:

  • As stated in this paper it is not sufficient to normalize dataset. This makes sense since the Bitcoin price range is from 0.06USD to 19,343USD. I think if data is fed to an ANN as a sliding window then each window has to be normalized.
  • Methods like Min-Max and Z-Score are good for pattern comparison but also standardize range. A pattern between 1000USD and 1200USD becomes exactly the same as same pattern between 1000USD and 1400USD so valuable information is lost. Of course it's possible to create an additional feature to represent volatility.

So what is the best way to normalize volatile time series datalike the price of Bitcoin?