Modified Z-Score

Modified Z-Score is a variation of the Z-score used for detecting outliers. It's particularly useful for time series data when the data doesn't necessarily follow a normal distribution. Unlike the traditional Z-score that uses the mean and standard deviation, the Modified Z-Score uses the median and Median Absolute Deviation (MAD) to avoid being skewed by extreme values.

Modified Z-Score Parameters

Param Name	Description	Default Values	Possible Values
Median	Median of the dataset	Computed from data	-
MAD	Median Absolute Deviation	Computed from data	-
z_thresh	Z-score threshold for outlier detection	3.5	2.5, 3.0, 3.5 (standard), 4.0+
Modified Z-Score	( 0.6745 \times \frac{(x_i - \text{median})}{\text{MAD}} )	Computed per value	-