Outlier Detection

IQR (Interquartile Range) Method

The IQR method is effective for detecting outliers in skewed or non-normal data. It uses the spread of the middle 50% of the data to determine outliers.

How it works:

1. Compute Q1 (25th percentile) and Q3 (75th percentile).
2. Calculate IQR = Q3 - Q1.
3. Flag points outside Q1 - 1.5 × IQR and Q3 + 1.5 × IQR as outliers.

Limitations:

• May not be suitable for small datasets where quartiles are unstable.
• Does not consider relationships in multivariate data.
• Assumes the data is roughly normally distributed (though it's more robust than some other methods).