Aug 03, 2022
In Design Forum
There is no big data field that does not do data analysis, and big data risk control is no exception. My point of view is that risk control and other Internet businesses are interoperable. This article introduces the data analysis that must be done in risk control to illustrate that data analysis is all-in-one. If a worker wants to do a good job, he must first sharpen his tools. Let’s talk about data analysis tools first. Analysis tools, the most common include Excel, SQL and Python. Even if everyone is in a technical position, there is no need for skill discrimination. Using Python is not more advanced than using Excel and SQL. Algorithm engineers laugh at themselves as SQL boy. SQL is the most used skill by data analysts and algorithm engineers. Whether the mobile number list correct data can be extracted from the massive business data is the premise of solving the problem . And Excel pivot tables are so powerful that everything can be pivoted. It's no exaggeration to say that I've never seen a problem that a pivot table can't solve. The data analysis platform, including metabase for open source and tableau for fee, can connect to the database for real-time interaction and provide rich smart dashboards. Personally recommend the open source BI tool Metabase. It has data analysis and query functions in a visual operation interface, allowing users who do not understand SQL to quickly grasp business data, and also support teams to share business data. It is a very good BI solution. 1. Business understanding If a financial institution hires you to consult on their risk management business, do you know what to do? Don't tell me you want to copy the wind control modeling competition. Don't drop the price. Solutions must be uniquely tailored to the current business and user base. You can graft on the experience of kaggle competitions, but stand on the shoulders of giants. For example, if you train a face recognition tool, you can't find the classic network results and everything will be fine, you need to go to fine-tune. So how do you understand business.