Challenges and opportunities in applying Neural Temporal Point Processes to large scale industry data

TL;DR

This paper explores the challenges of applying Neural Temporal Point Processes to large-scale industry data, highlighting issues like data imbalance, scalability, and proposing solutions such as static feature integration.

Contribution

It systematically reproduces existing NTPP models on benchmarks and applies them to a larger real-world dataset, identifying key limitations and proposing a static feature parametrization.

Findings

NTPP models struggle with rare event forecasting.

Stochastic differential equation-based NTPPs do not scale well.

Static user features help mitigate cold-start problems.

Abstract

In this work, we identify open research opportunities in applying Neural Temporal Point Process (NTPP) models to industry scale customer behavior data by carefully reproducing NTPP models published up to date on known literature benchmarks as well as applying NTPP models to a novel, real world consumer behavior dataset that is twice as large as the largest publicly available NTPP benchmark. We identify the following challenges. First, NTPP models, albeit their generative nature, remain vulnerable to dataset imbalances and cannot forecast rare events. Second, NTPP models based on stochastic differential equations, despite their theoretical appeal and leading performance on literature benchmarks, do not scale easily to large industry-scale data. The former is in light of previously made observations on deep generative models. Additionally, to combat a cold-start problem, we explore a novel addition to NTPP models - a parametrization based on static user features.