Outlier values be treated by:
1. Imputing or assigning a new value. One can simply replace the outlier with the median or mean of the variable, or alternatively use a regression model to predict the missing value.
2. Transforming the given data set, for example logarithmic transformation, square root transformation, or square transformation, percentile transformation, and then use the new transformed field instead of the original field.
3. Capping (or flooring) the outliers data at a level that keeps it intact.
4. Trimming involves dropping or eliminating the outlier data completely, to prevent it from skewing the analysis.
However, the method to apply depends on factors such as dataset size, how far the outlier is out of line with the rest of the dataset.