Risk Intelligence: A Digital Risk Assessment System Based on SFI Bitwise Fuzzification

For a long time, risk management was a rather abstract subject to me, and I did not take it seriously. That changed when I met a respected colleague and my research advisor, Oleg M., editor-in-chief of the publication “Risk Management in Agribusiness.” Over years of joint research into agricultural risks, risk stopped being an abstraction and became an understandable mechanism. I owe Oleg a great deal for that.

My idea is to develop a universal way of assessing risks using discrete data packed into bit vectors. My hypothesis is that risk assessment does not require exact numerical values. What it needs is a very large number of different kinds of reliable discrete data. The more diverse, the better.

I have built a prototype and am debugging it. Overall, the approach looks promising, but the risk estimates need calibration against real-world data, which takes considerable time. Still, a start has been made. More to come.