There are two types of noise in time series data, white noise, and red noise.

**White Noise:**

A noise sequence is a white noise sequence if

- The expectation of each element is 0.
- The variance of each element is ﬁnite.
- The elements are uncorrelated.

To check white noise Fisher’s test is performed. Fisher’s test checks that if a sequence is white noise sequence or not. If the value of Fisher’s test outcome is less than 0.5, the sequence is considered as noise free.

**Red Noise:**

A time series is a red noise sequence if

- The sequence has zero mean.
- The sequence has constant variance.
- The Sequence has a serial correlation in time.

Red noise has a power spectrum towards low frequencies.