Hengxu Lin (Pronunciation)

Understanding and predicting consumer preferences in dynamic markets is crucial for marketing decision-making. Existing methods for modeling dynamic consumer heterogeneity are either inflexible in representing individual-level preference trajectories or as in the case of recently proposed Gaussian process-based approaches, induce unimodal cross-sectional heterogeneity distributions. This prevents them from accurately capturing the growth, decline, or divergence of consumer segments over time. We introduce a novel Bayesian nonparametric machine learning framework that leverages Dependent Dirichlet Processes for modeling hetero- geneity to overcome these methodological constraints. Our approach combines the strengths of Dirichlet processes and Gaussian processes to flexibly model a collection of temporally corre- lated population distributions. This structure simultaneously captures smooth individual-level preference trajectories and evolving multimodal population distributions. Extensive simulations show that our model significantly outperforms a state-of-the-art GP-based benchmark in recovering parameters and predicting choices when the evolving population distributions are multimodal and performs comparably when they are unimodal. An application to IRI scanner data for multiple categories spanning the Great Recession shows that our model provides more plausible price elasticity estimates, improved predictions, and valuable substantive insights into recession-induced shifts in consumer behavior. We finally show that coupon targeting strategies based on our model can generate superior profits.

We develop a reinforcement learning model that maximizes the long-run conversion rate for Expedia hotel bookings. The initial state of the system is characterized by a combination of any previous information Expedia has about a customer and the information he/she provides to initiate a hotel search. Expedia’s action takes the form of presenting an ordered subset of the relevant hotels. The customer’s response, such as examining some hotels for more detail and/or changing the filters, transitions the system to a new state. Expedia can use the explicit and implicit information it has obtained from the preceding customer interactions to display another ranked subset of hotels. The process stops when the customer leaves the website with or without making a hotel booking. The values of all Expedia actions are assessed only by their effect on the consumer’s purchase decision. The number of possible actions and states can be extremely large, making it impossible to estimate the state and action value functions before choosing optimal actions. Reinforcement learning uses trial-and-error to explore promising actions, and temporal-difference learning to update value functions. We combine the actor-critic framework for reinforcement learning with a deep-learning model in which the state and action values are functions of contextual variables.

Modeling and managing portfolio risk is perhaps the most important step to achieve growing and preserving investment performance. Within the modern portfolio construction framework that built on Markowitz's theory, the covariance matrix of stock returns is required to model the portfolio risk. Traditional approaches to estimate the covariance matrix are based on human designed risk factors, which often requires tremendous time and effort to design better risk factors to improve the covariance estimation. In this work, we formulate the quest of mining risk factors as a learning problem and propose a deep learning solution to effectively "design" risk factors with neural networks. The learning objective is carefully set to ensure the learned risk factors are effective in explaining stock returns as well as have desired orthogonality and stability. Our experiments on the stock market data demonstrate the effectiveness of the proposed method: our method can obtain 1.9% higher explained variance measured by R2 and also reduce the risk of a global minimum variance portfolio. Incremental analysis further supports our design of both the architecture and the learning objective.

GitHub

Microsoft Research Press

Successful quantitative investment usually relies on precise predictions of the future movement of the stock price. Recently, machine learning based solutions have shown their capacity to give more accurate stock prediction and become indispensable components in modern quantitative investment systems. However, the i.i.d. assumption behind existing methods is inconsistent with the existence of diverse trading patterns1 in the stock market, which inevitably limits their ability to achieve better stock prediction performance. In this paper, we propose a novel architecture, Temporal Routing Adaptor (TRA), to empower existing stock prediction models with the ability to model multiple stock trading patterns. Essentially, TRA is a lightweight module that consists of a set of independent predictors for learning multiple patterns as well as a router to dispatch samples to different predictors. Nevertheless, the lack of explicit pattern identifiers makes it quite challenging to train an effective TRA-based model. To tackle this challenge, we further design a learning algorithm based on Optimal Transport (OT) to obtain the optimal sample to predictor assignment and effectively optimize the router with such assignment through an auxiliary loss term. Experiments on the real-world stock ranking task show that compared to the state-of-the-art baselines, e.g., Attention LSTM and Transformer, the proposed method can improve information coefficient (IC) from 0.053 to 0.059 and 0.051 to 0.056 respectively. Our dataset and code used in this work are publicly available.

GitHub

KDD21 Accepted Version and Video
Poster
Slides
Microsoft Research Press