Z-Score Filtering

Z-score filtering is a statistical technique used to identify and remove outliers by measuring how many standard deviations a data point is from the mean. In financial data, a high Z-score indicates that a price or volume change is statistically unlikely, suggesting it may be an error or a transient anomaly.

By setting a threshold, analysts can automatically exclude these extreme values from their datasets. This is particularly useful for cleaning noisy crypto data, where extreme spikes are common but often unrepresentative of true market conditions.

Z-score filtering is a simple yet powerful tool for maintaining the quality of quantitative models. It helps ensure that volatility estimates and risk metrics are not distorted by outliers.

However, it must be used carefully, as genuine market events could also be filtered out if the threshold is set too strictly. It is a foundational tool in the data cleaning pipeline.