Claims:
1. A computer implemented method, comprising:
(a) processing a time series data to obtain one or more blocks of observation values;
(b) reading a first observation value from said one or more blocks of observation values;
(c) updating a set of convolution values stored in a memory with said first observation value;
(d) updating a partial sum by adding said first observation value to said partial sum;
(e) incrementing a count of observations in a count variable each time an observation value is read;
(f) repeating (b) till (e) until a last observation value is read from a last block from said one or more blocks of observation values to obtain an updated set of convolution values, an updated partial sum, and an updated count;
(g) computing a first set of elements for a first matrix using at least one of (i) said updated set of convolutions values, (ii) summation of observation values computed from said updated partial sum, or (iii) said updated count;
(h) computing a second set of elements for a second matrix using at least one of (i) said updated set of convolution values, or (ii) said summation of observation values computed from said partial sum; and
(i) generating a vector of autoregression coefficients based on said first matrix and said second matrix.

2. The computer implemented method of claim 1, wherein said time series data is obtained from one or more sensors.

3. The computer implemented method of claim 1, wherein said vector of autoregression coefficients is generated in a batch mode.

4. The computer implemented method of claim 1, wherein said vector of autoregression coefficients is generated at one or more pre-defined intervals in a streaming mode.

5. A computer implemented system, comprising:
a memory storing instructions, and a set of convolution values;
a communication interface; and
a hardware processor coupled to said memory, wherein said hardware processor is configured by said instructions to
(a) process a time series data to obtain one or more blocks of observation values,
(b) read a first observation value from said one or more blocks of observation values,
(c) update a set of convolution values stored in said memory with said first observation value,
(d) update a partial sum by adding said first observation value to said partial sum,
(e) increment a count of observations in a count variable each time an observation value is read,
(f) repeat (b) till (e) until a last observation value is read from a last block from said one or more blocks of observation values to obtain an updated set of convolution values, an updated partial sum, and an updated count,
(g) compute a first set of elements for a first matrix using at least one of (i) said updated set of convolutions values, (ii) summation of observation values computed from said updated partial sum, or (iii) said updated count,
(h) compute a second set of elements for a second matrix using at least one of (i) said updated set of convolution values, or (ii) said summation of observation values computed from said partial sum, and
(i) generate a vector of autoregression coefficients based on said first matrix and said second matrix.

6. The computer implemented system of claim 5, wherein said time series data is obtained from one or more sensors.

7. The computer implemented system of claim 5, wherein said vector of autoregression coefficients is generated in a batch mode.

8. The computer implemented system of claim 5, wherein said vector of autoregression coefficients is generated at one or more pre-defined intervals in a streaming mode.
, Description:As Attached