How can I detect and handle outliers in a dataset?

My dataset contains outliers that may affect model performance. What approaches and techniques can be used to identify and appropriately handle outliers in data preprocessing for machine learning tasks?