# 36th place Solution

I recently felt like I was going through a slump, and I didn't have high expectations, but fortunately, I managed to secure 36th place.
Here are my three solutions:
NN Pairwise Ranking
Classifier + CAT + XGB
XGBoost with Monotonicity
You can check out the solutions in the following notebooks.
[https://www.kaggle.com/code/junhanzangai/cibmtr-nn-lgb-cat-ensemble-of-solutions](https://www.kaggle.com/code/junhanzangai/cibmtr-nn-lgb-cat-ensemble-of-solutions)
Additionally, using MinMaxScaler to normalize each prediction to a 0–1 range and making predictions based on this normalization was a crucial factor. You can see this clearly in the following notebook.
I believe my score could have improved further, if I incorporated DeepResidualNN into my approach.
![](assets/07e0b5b26102ceb3b29331c6.png)