CS488_Project

This is a course project for Point Process and Its Applications(CS488), ACM Class 2017, SJTU. Written in 2020.

Introduction

With the development of deep learning, the neural temporal point process is becoming more and more popular. Neural point process model is good at prediction, while it is weak on interpretability. The project is to build a Interpretable Neural Temporal Point Process model (INTPP) based on the paper "Improving Interpretability and Predictive Performance of Neural Temporal Point Processes", which is good at prediction and interpretability simultaneously. I also conduct experiments on both synthetic sequences which are generated in assignment1 and ATMs dataset.

For more details, please refer to my report.