Ensemble Pricing Model for Data Assets with Ranking-Pruning-averaging Strategy

Data has become an important strategic resource and factor of production. Data pricing is the core issue of data trading and is the pivot to form a virtuous cycle mechanism of data value from generation, multiplication to empowerment. Since it is difficult for a single pricing model to comprehensively capture the overall patterns with data asset prices, this paper establishes a novel ensemble pricing model for data assets. The effective ensemble of pricing results is achieved by generating ten machine learning pricing models and introducing a ranking pruning average strategy. Finally, the study conducts numerical experiments with real data from data trading platforms to verify the superiority of the pricing model, showing that the ensemble pricing model can be regarded as a promising tool for data pricing.