Method
Stacking
Meta-model trained on base-model predictions.
9 cases from 8 companies apply Stacking. 4 report results in production.
Typical tasks: Forecast demand and load, Optimize decisions, Catch fraud and abuse, Moderate content at scale and 1 more. Companies: Walmart, Wildberries, Stripe, DoorDash, Alfa-Bank and 3 more. Often combined with: Gradient boosting, Prophet, Logistic regression, Lag and window features, Backtesting and 33 more.
| Tabular Neural Networks for Combining Heterogeneous Banking Data: Challenges and Limitations | — | banking | 2026-05 |
| Autotuning: A comprehensive approach to Time Series Forecasting | retail | 2025-03 | |
Developing a Recommender System for Ammonia Production![]() | ![]() | manufacturing | 2024-05 |
How We Fought Spam in Reviews![]() | ![]() | marketplace | 2024-05 |
| How we built it: Smart Retries | fintech | 2024-01 | |
| Order Forecasting on a Map Using a Single Time Series | general | 2023-10 | |
| Through Thorns to Data: Challenges and Successes in Searching and Processing Transactions in FinTech | banking | 2023-10 | |
| How DoorDash Built an Ensemble Learning Model for Time Series Forecasting | food-delivery | 2023-06 | |
| Neural Network Models on Card Transactions | banking | 2022-06 |

