I’m currently in a condition where i met this kind of outlier visualization, which when i think about it i have been encountering it for quite a few times with different datasets .

When faced with this kind of problem, what action should i take. Clearing it out would just makes the dataset size become small as other attributes also have similar outliers. I thought of fill it with another value, but would it affect my model accuracy?