Abstract
Inventory backorder prediction is widely recognized as an important component of inventory models. However, backorder prediction is traditionally based on stochastic approximation, thus neglecting the substantial amount of useful information hidden in historical inventory data. To provide those inventory models with a big data-driven backorder prediction, we propose a machine learning model equipped with an undersampling procedure to maximize the expected profit of backorder decisions. This is achieved by integrating the proposed profit-based measure into the prediction model and optimizing the decision threshold to identify the optimal backorder strategy. We show that the proposed inventory backorder prediction model shows better prediction and profit function performance than the state-of-the-art machine learning methods used for large imbalanced data. Notably, the proposed model is computationally effective and robust to variation in both warehousing/inventory cost and sales margin. In addition, the model predicts both major (non-backorder items) and minor (backorder items) classes in a benchmark dataset.
Highlights
In customers’ purchasing pattern forecasting, it is discovered that consumers favor their demands to be backordered when inventory goes in shortfalls
This study aims to investigate the question of how a profit function-maximizing inventory backorder prediction system provides quantitative insights into the economic merit of optimal backorder strategy, in the era of big data
This paper finds that the material backorder prediction based on big data characteristics and its profitability occurring from misclassification have not been considered in previous studies
Summary
In customers’ purchasing pattern forecasting, it is discovered that consumers favor their demands to be backordered when inventory goes in shortfalls. DATA This study validates the profit function-maximizing inventory management policy by applying a real-world material backorder big dataset. It comes from the Kaggle’s contest, namely Can You Predict Product Backorders? The dataset includes the following attributes: item identification (stock keeping unit, sku; used to track inventory levels), present inventory size (national_inv), registered transit time (lead_time), item quantity in transit (in_transit_qty), sales forecasts (forecast_3_month, forecast_6_month, forecast_9_month), prior sales volumes (sales_1_month, sales_3_month, sales_6_ month, sales_9_month), minimum amount of stock recommended (min_back), source issue for item identified (potential_issue), amount of overdue from source (pieces_past_due), prior source performance (perf_6_month_ avg, perf_12_month_avg), amount of stock overdue (local_bo_qty), risk flags (deck_risk, operating entities constraint oe_constraint, production part approval process risk ppap_risk, stop_auto_buy, rev_stop) and a response variable (went_on_backorder). The dataset was pre-processed through attribute standardization and missing value imputation using the fuzzy k-means algorithm, as recommended by [58] under general assumptions
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
