Demand Prediction on H&M Personalized Fashion Recommendations (10 random train-test splits)
0.94CRPSLLM-cal
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| LLM-calModel type=logit-calibrated version of mixture-of-persona model, Reference section=2.22026.06 | 0.94 | 0.3 | 1.38 | 1.79 | |
| embModel type=embedding-based model, Reference section=3.12026.06 | 0.99 | 0.32 | 1.44 | 1.86 | |
| LLMModel type=mixture-of-persona model, Reference section=2.12026.06 | 1 | 0.32 | 1.44 | 1.87 | |
| normalModel type=normally distributed model, Reference section=3.22026.06 | 1.19 | 0.43 | 1.46 | 1.86 |