The aim of current application is to work on Machine learning methodologies on space-time predictive analysis in FinTech. We focus on structuring the research topic and extending the methodology to exploit the spatial and time-series variations of the banking sector during the pandemic of Covid-19 via means of Machine Learning.
The methodology that we have already applied for this particular use case consists of exploratory data analysis, simple linear models for each month with spatial diagnosis, and spatial panel data models with fixed effects. The results so far are statistically significant and in favour of a positive relationship between the variables of interest. Increasing Covid-19 infections result into higher loan and saving rates, with significant spatial interactions given by the neighbouring counties. As the spatial and temporal effects are in yielding better models, our goal is to apply more novel methods on top of the dataset to account for these effects.