Causality Learning From Time Series Data For The Industrial Finance Analysis Via The Multi-Dimensional Point Process

Causality learning has been an important tool for decision making, especially for financial analytics. Given the time series data, most existing works construct the causality network with the traditional regression models and estimate the causality by pairs. To fulfil a holistic one-shot inference procedure over the whole network, we propose a new causal inference method for the multi-dimensional time series data, specifically related to some case studies for the industrial finance analytics. Specifically, the time series are first converted to the event sequences with timestamps by fluctuation the detection, and then a multi-dimensional point process is used for learning the underlying causality among the event sequences, which we assume stands for the relations among the time series. The expectation-maximization algorithm is used for minimizing the negative log-likelihood with the regularization in order to avoid overfitting in the high dimension and will make the causal inference more reasonable. Over 250 factors with time series data related to the industrial finance are used in this paper to evaluate the model and the experimental showcase of the superiority of our approach on the real-world finance data.