Hierarchical Dynamic Modeling for Individualized Bayesian Forecasting

We present a case study and methodological developments in large-scale hierarchical dynamic modeling for personalized prediction in commerce. The context is supermarket sales, where improved forecasting of customer/household-specific purchasing behavior informs decisions about personalized pricing and promotions on a continuing basis. This is a big data, big modeling and forecasting setting involving many thousands of customers and items on sale, requiring sequential analysis, addressing information flows at multiple levels over time, and with heterogeneity of customer profiles and item categories. Models developed are fully Bayesian, interpretable and multi-scale, with hierarchical forms overlaid on the inherent structure of the retail setting. Customer behavior is modeled at several levels of aggregation, and information flows from aggregate to individual levels. Forecasting at an individual household level infers price sensitivity to inform personalized pricing and promotion decisions. Methodological innovations include extensions of Bayesian dynamic mixture models, their integration into multi-scale systems, and forecast evaluation with context-specific metrics. The use of simultaneous predictors from multiple hierarchical levels improves forecasts at the customer-item level of main interest. This is evidenced across many different households and items, indicating the utility of the modeling framework for this and other individualized forecasting applications.

[1]  M. West,et al.  Embracing the complexity of genomic data for personalized medicine. , 2006, Genome research.

[2]  Lindsay Rebecca Berry.  Bayesian Dynamic Modeling and Forecasting of Count Time Series , 2019 .

[3]  Mike West,et al.  Bayesian Computation in Dynamic Latent Factor Models , 2020, Journal of Computational and Graphical Statistics.

[4]  Lars Schmidt-Thieme,et al.  BPR: Bayesian Personalized Ranking from Implicit Feedback , 2009, UAI.

[5]  Xiaohui Hu,et al.  Web Service Recommendation Based on Time Series Forecasting and Collaborative Filtering , 2015, 2015 IEEE International Conference on Web Services.

[6]  Tim Januschowski,et al.  Deep Factors for Forecasting , 2019, ICML.

[7]  Holly Dressman,et al.  Towards integrated clinico-genomic models for personalized medicine: combining gene expression signatures and clinical factors in breast cancer outcomes prediction. , 2003, Human molecular genetics.

[8]  Hsiang-Fu Yu,et al.  Think Globally, Act Locally: A Deep Neural Network Approach to High-Dimensional Time Series Forecasting , 2019, NeurIPS.

[9]  Lars Schmidt-Thieme,et al.  Personalized Forecasting Student Performance , 2011, 2011 IEEE 11th International Conference on Advanced Learning Technologies.

[10]  Ghassen Jerfel,et al.  Dynamic Collaborative Filtering With Compound Poisson Factorization , 2016, AISTATS.

[11]  Michael Bohlke-Schneider,et al.  High-Dimensional Multivariate Forecasting with Low-Rank Gaussian Copula Processes , 2019, NeurIPS.

[12]  Ute Beyer,et al.  Bayesian Forecasting And Dynamic Models , 2016 .

[13]  Brian Keng,et al.  Multivariate Arrival Times with Recurrent Neural Networks for Personalized Demand Forecasting , 2018, 2018 IEEE International Conference on Data Mining Workshops (ICDMW).

[14]  Jonathan R. Walsh,et al.  Machine learning for comprehensive forecasting of Alzheimer’s Disease progression , 2018, Scientific Reports.

[15]  Mike West,et al.  Bayesian Decision Analysis and Constrained Forecasting , 2020 .

[16]  David Hsu,et al.  Exploration in Interactive Personalized Music Recommendation: A Reinforcement Learning Approach , 2013, TOMM.

[17]  Yinghai Lu,et al.  Deep Learning Recommendation Model for Personalization and Recommendation Systems , 2019, ArXiv.

[18]  Wei Chu,et al.  Personalized recommendation on dynamic content using predictive bilinear models , 2009, WWW '09.

[19]  Hanning Zhou,et al.  Collaborative Filtering with User-Item Co-Autoregressive Models , 2016, AAAI.

[20]  Taghi M. Khoshgoftaar,et al.  A Survey of Collaborative Filtering Techniques , 2009, Adv. Artif. Intell..

[21]  Mike West,et al.  Bayesian Forecasting of Many Count-Valued Time Series , 2018, Journal of Business & Economic Statistics.

[22]  Padhraic Smyth,et al.  Prediction of Sparse User-Item Consumption Rates with Zero-Inflated Poisson Regression , 2018, WWW.

[23]  M. West,et al.  Integrated modeling of clinical and gene expression information for personalized prediction of disease outcomes. , 2004, Proceedings of the National Academy of Sciences of the United States of America.

[24]  Ognjen Rudovic,et al.  Personalized Gaussian Processes for Future Prediction of Alzheimer's Disease Progression , 2017, ArXiv.

[25]  Wei Niu,et al.  Neural Personalized Ranking for Image Recommendation , 2018, WSDM.

[26]  Valentin Flunkert,et al.  DeepAR: Probabilistic Forecasting with Autoregressive Recurrent Networks , 2017, International Journal of Forecasting.

[27]  Xiaoyu Du,et al.  Adversarial Personalized Ranking for Recommendation , 2018, SIGIR.

[28]  Mike West,et al.  Probabilistic forecasting of heterogeneous consumer transaction–sales time series , 2018, International Journal of Forecasting.

[29]  Pooyan Kazemian,et al.  Personalized Prediction of Glaucoma Progression Under Different Target Intraocular Pressure Levels Using Filtered Forecasting Methods. , 2017, Ophthalmology.