|
Abstract:
To explore the inherent law of financial market price fluctuation and the characteristics of financial time series in the age of mass data, which is significant to understand and predict the trend of financial market. Nowadays, data becomes have more various types and large volume, under the background of massive data, we find that: the traditional long memory test method's computation speed cannot meet the requirement of modern long memory test research. The paper takes V/S method as an example, in order to compressing the sample size , reducing computational complexity and simplifying rapid calculations, it adopts the characteristic points of extremum. Empirical findings: the results of V/S long memory test are nearly unanimous between the fast detection method and the original one, especially for k value is 1, meanwhile, it could greatly improved time efficiency under guarantee the accuracy of this test.
|
