Depression Episodes Detection in Unipolar and Bipolar Patients: A Methodology with Feature Extraction and Feature Selection with Genetic Algorithms Using Activity Motion Signal as Information Source

Table 1

Statistical features extracted from the temporal and frequency data.

Feature	Description
Mean
Standard deviation
Variance
Trimmed mean	Mean with outliers trimmed
Coefficient of variation
Inverse coefficient of variation
Kurtosis
Skewness
Quantiles: 1, 5, 25, 75, 95, 99%

represents the median value; 1  i  9; (j − m)/n  p  (j − m + 1)/n; represents the order statistic; n represents the sample size; γ is in function of j and , where and ; and m represents a constant determined by the sample quantile type.