Deep learning improves demand forecasts despite sales data limits
A Deep Learning Framework for Quantile Demand Forecasting in E-commerce with Doubly Censored Data
Machine Learning
Summary
Predicting how much stuff people will buy online is hard because actual sales numbers don’t always show true demand—sales might be limited by low stock or boosted by sales promotions. The authors created a deep learning method that estimates ranges of possible demand even when the sales data is incomplete or unclear. Their approach works better than other methods and helps stores plan inventory and sales strategies more reliably. This means better decisions for online sellers who want to avoid running out of products or having too much leftover stock.
What this means in practice
- •For online retail operations teams: Improve inventory planning by forecasting demand more accurately when sales data is incomplete due to stock limits or promotions.
- •For supply chain planners: Use the method’s improved uncertainty estimates to reduce overstock and stockouts with better demand predictions under ambiguous sales conditions.
Authors
Ziwen Ye, Jiawen Wei, Teng Song, Chen Chen, Guangrui Ma
Abstract
In recent years, a growing number of online retail companies have embraced deep learning models for demand forecasting, leveraging their scalability in processing large-scale datasets. Accurate demand forecasts enable improved planning and allow operational flexibility to adjust sales accordingly. Moreover, quantile estimation has gained widespread application as it offers valuable uncertainty quantification, thereby enhancing the robustness of decision-making processes. However, limited research has addressed the effective quantification of uncertainty when target labels are ambiguous. In retail, actual sales can be capped by stock and boosted by temporary promotions, whereas the target is underlying demand. Such scenarios are characterized by double censoring and no deep-learning pipeline yet exists for such data. In this paper, we introduce a straightforward framework for learning quantiles from doubly-censored observations and propose a tailored evaluation metric. Our approach is supported by theoretical guarantees, and we empirically demonstrate that our metric correlates well with the oracle calibration measure. Experimental results show that our estimator surpasses existing baselines. Finally, inventory simulations and real-world applications confirm the practical utility of our method in retail operations.