发明名称 |
Heteroscedastic Data Compression Using Arima-Garch Model Estimation |
摘要 |
Methods and apparatus are provided for compression and decompression of heteroscedastic data, such as seismic data, using Autoregressive Integrated Moving Average (ARIMA)-Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model estimation. Heteroscedastic data is compressed by obtaining the heteroscedastic data; applying the heteroscedastic data to an ARIMA-GARCH model; determining residuals between the obtained heteroscedastic data and the ARIMA-GARCH model; and compressing parameters of the ARIMA-GARCH model and the residuals using entropy encoding, such as an arithmetic encoding, to generate compressed residual data. Parameters of the ARIMA-GARCH model are adapted to fit the obtained heteroscedastic data. The compressed residual data is decompressed by performing an entropy decoding and obtaining the parameters of the ARIMA-GARCH model and the residuals. The ARIMA-GARCH model predicts heteroscedastic data values and then the decompressed residuals are added. |
申请公布号 |
US2015278284(A1) |
申请公布日期 |
2015.10.01 |
申请号 |
US201414230510 |
申请日期 |
2014.03.31 |
申请人 |
EMC Corporation |
发明人 |
Bordignon Alex L.;Ciarlini Angelo E.M.;Voyt Timothy A.;Rossetto Silvana;Medeiros Renato M.M. |
分类号 |
G06F17/30 |
主分类号 |
G06F17/30 |
代理机构 |
|
代理人 |
|
主权项 |
1. A method for compressing heteroscedastic data, said method comprising:
obtaining said heteroscedastic data; applying said heteroscedastic data to an Autoregressive Integrated Moving Average (ARIMA)-Generalized Autoregressive Conditional Heteroscedasticity (GARCH) model; determining residuals between said obtained heteroscedastic data and said ARIMA-GARCH model; and compressing parameters of said ARIMA-GARCH model and said residuals using entropy encoding to generate compressed residual data. |
地址 |
Hopkinton MA US |