[doc] Update competition winning solutions (#12087)

Add three verified 1st-place wins where XGBoost was a primary model:

- ACM RecSys Challenge 2021 (Schifferer, Deotte, Titericz, Liu)
- Kaggle Playground Series S5E2 Backpack Prediction (Deotte, Feb 2025)
- March Machine Learning Mania 2025 (Odeh, $50K prize)

Author: PavelGuzenfeld

demo/README.md

@@ -91,6 +91,9 @@ Please send pull requests if you find ones that are missing here.
 - David Austin, Weimin Wang, 1st place winner in [Iceberg-classifier-challenge](https://www.kaggle.com/c/statoil-iceberg-classifier-challenge/leaderboard) Link to [discussion](https://www.kaggle.com/c/statoil-iceberg-classifier-challenge/discussion/48241)
 - Kazuki Onodera, Kazuki Fujikawa, 2nd place winner in [OpenVaccine: COVID-19 mRNA Vaccine Degradation Prediction](https://www.kaggle.com/c/stanford-covid-vaccine/overview) Link to [Discussion](https://www.kaggle.com/c/stanford-covid-vaccine/discussion/189709)
 - Prarthana Bhat, 2nd place winner in [DYD Competition](https://datahack.analyticsvidhya.com/contest/date-your-data/). Link to [Solution](https://github.com/analyticsvidhya/DateYourData/blob/master/Prathna_Bhat_Model.R).
+- Benedikt Schifferer, Chris Deotte, Gilberto Titericz, Bo Liu, 1st place winner of [ACM RecSys Challenge 2021](https://recsys.acm.org/recsys21/challenge/). An ensemble of 5 XGBoost models and 3 neural networks. Link to [writeup](https://medium.com/nvidia-merlin/winning-the-recsys2021-challenge-by-a-diverse-set-of-xgboost-and-neural-network-models-4c5422a642d8) and [paper](https://dl.acm.org/doi/10.1145/3487572.3487605).
+- Chris Deotte, 1st place winner of [Kaggle Playground Series S5E2 - Backpack Prediction Challenge](https://www.kaggle.com/competitions/playground-series-s5e2) (Feb 2025). A single XGBoost model with 500 engineered features. Link to [discussion](https://www.kaggle.com/competitions/playground-series-s5e2/discussion/565539).
+- Mohammad Odeh, 1st place winner of [March Machine Learning Mania 2025](https://www.kaggle.com/competitions/march-machine-learning-mania-2025) ($50K prize). XGBoost outperformed CatBoost and LightGBM. Link to [writeup](https://www.kaggle.com/competitions/march-machine-learning-mania-2025/writeups/mohammad-odeh-first-place-solution).
 
 ## Talks
 - XGBoost: A Scalable Tree Boosting System ([video] (https://www.youtube.com/watch?v=Vly8xGnNiWs) + [slides](https://speakerdeck.com/datasciencela/tianqi-chen-xgboost-overview-and-latest-news-la-meetup-talk)) by Tianqi Chen at the Los Angeles Data Science meetup