2.1.3. Data Pre-Processing

First, a random seed, as 91,190,530, was defined for replicability purposes. Then, five observations that had abnormal values were removed. In particular, one observation had abnormally high values, while other four were filled with the same exact dummy value for all adjectives. None of these observations add the *selected\_attr* feature filled. The final dataset is made of 250 observations, with the next lines describing the entire treatment and all applied methods, including synthetic data creation.
