Company
Target
11 cases from Target (Retail, E-commerce), each digested from a public talk or engineering post and linked to the original. 6 were checked in an online A/B test; 4 report results in production.
Tasks: Optimize decisions, Forecast demand and load, Balance relevance and diversity, Rank search results, Recommend from what the user just did and 2 more. Methods: Multi-armed bandit, Graph neural network, BiLSTM, Focal loss, A/B test, Greedy algorithm and 22 more. Tools: Apache Spark, TensorFlow, Kubernetes, CatBoost, LightGBM, XGBoost and 4 more.
Tech blog: target.github.io