Prompt-engineered embeddings fail to beat strong tabular models in demand forecasting benchmark
A new benchmark finds that prompt-engineered contextual embeddings improved Random Forest but failed to outperform a strong tabular XGBoost model ...
Tag: time-series analytics
