How to measure Covid impact by analysing credit card transaction of customer

I Want to know how can I identify that is the customer is in financial distress due to the COVID situation using its credit card transactions.

I have a daily transaction of customers till current date, Any thoughts or ideas would be really helpful.

For example, my thoughts that if sudden increment in credit card utilization then it can be an indicator that person in financial distress as could be a flag to potential lenders or creditors that having trouble managing your finances.

Topic data-science-model data-analysis finance statistics predictive-modeling

Category Data Science

1
answered at

Do you have some sort of labeled data?

Otherwise I'd hypothesize that this task is close to impossible since any sort of unsupervised algorithm using anomaly detection would likely have an incredibly low precision since almost all data would be classified as an anomaly during this time.

However, if you were to obtain some labeled data (labeling customers as 'in financial distress' and 'not in financial distress') then you could use some supervised learning algorithm to try to predict whether the unlabeled customers fall in category A or category B.

Unless you acquire some sort of domain expertise or labeled data, this would be a very hard task.

About

Geeks Mental is a community that publishes articles and tutorials about Web, Android, Data Science, new techniques and Linux security.