In the era of big data, new data formats and complex relationships between variables bring unprecedented challenges to econometrics. However, on the other hand, it also presents a unique opportunity for the development and original theoretical breakthroughs in econometrics. Specifically, RIDE will focus on the following two research directions:

First, modeling and application of high-dimensional unstructured data. This is because when dealing with big data, statistical and econometric models face two major challenges. Firstly, over 80% of data is unstructured (meaning it doesn't have specific data models to describe it and is random and vague). Secondly, data is high-dimensional and nonlinear. In large datasets, the technical and economic value of individual data points is very low. To extract and mine useful information and structure from big data, at least a subset of the dataset must be used, making this problem a high-dimensional (e.g., 500-5000 dimensions) and nonlinear data modeling issue. How to address these issues from a statistical theory perspective is of vital importance.

Second, modeling and application of spatiotemporal data. With the development of socioeconomics, data's social network and spatial-related characteristics are becoming stronger. In various fields such as economics, finance, management, and sociology, besides the correlation between variables, there is a general interdependence among different samples of the same variable, known as social networks or spatial dependence. Effectively utilizing this information in a big data environment is a crucial research direction for RIDE's future, and it's also one of the important development directions in current econometrics.