admin管理员组

文章数量:1122832

I have a data frame containing gene-TF pairs with two columns (colA and colB) showing the strength of the interaction. I have another column (colC) with continuous variables where I use it as a sanity check, meaning that higher the value in colC, that gene-gene pair should be ranked higher. Based on that, I am trying to rank these pairs.

  1. I thought XGBRanker would be ideal in this example. When it comes to grouping, I want model to compare each pair with each pair. Since the model requires grouping, I grouped pairs based on unique TF names. In total I had 25-30 groups, I dont think that it is ideal. For the rank column, I split the continuous variable ColC into quartiles and have 3-5 splits.

I am a bit confused since I couldnt find much documentation about XGBRanker. What would be the ideal approach to take it here?

Appreciate any help! Thanks

I ran the model with default parameters and basic grouping. First ranking didnt make so much sense. I am trying to figure out a better way of performing this.

本文标签： xgboostXGBRanker grouping and ranking specificsStack Overflow

版权声明：本文标题：xgboost - XGBRanker grouping and ranking specifics - Stack Overflow 内容由网友自发贡献，该文观点仅代表作者本人， 转载请联系作者并注明出处：http://www.betaflare.com/web/1736306626a1933025.html， 本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌抄袭侵权/违法违规的内容，一经查实，本站将立刻删除。