I'm dealing with binary option, I'm making a classification model using xgboost, my idea is to predict the posterior candle color by using some data of the 20 previous candles so I made a dataset on each column is the i-th candle I'd like that my xgboost give more importance to the last 5 previous candles than the 15 before (summing 20 candles). By more importance I mean the last 5 candles will be more informative to response than the last 15. How would I do that?