Тетерина Дарья Владимировна
Methods of Machine Learning for Censored Demand Prediction
In this paper, we analyze a new approach for demand prediction in retail taking into account data censorship and using machine learning methods. One of the significant gaps in demand prediction by machine learning methods is the unaccounted data censorship. Econometric approaches to modeling censored demand are used to obtain consistent and unbiased estimates of parameters. These approaches can also be transferred to different classes of machine learning models to reduce the prediction error of future prices or sales volumes. In this study we build two ensemble demand models with and without demand censorship, aggregating predictions for machine learning methods such as Ridge regression, LASSO and Random Forest. Having estimated the predictive properties of both models, we empirically prove the best predictive power of the model, taking into account the censored nature of demand.
Текст работы (работа добавлена 18 мая 2018г.)