Friday 15 October 2021

AI

Quant is (ML/NN) Ai or at least Complex statistics when used for trading equity (and other instruments) by various methods. 


While it's all AI for price forecasting or HFT and taking decisions to even take trades itself. 


Some algo's failed badly while some adapted to the Black Swan event (financial market's disaster) making even that to be profitable event


Quant Bust 2020 https://arxiv.org/abs/2006.05632 


During the same black Swan event, various decision makers (doctors) overused CT scans for screening covid and overused various drugs for prophylaxis and treatment  while various guidelines and minimalist doctor's choose to go with symptomatology and conservative with home care unless needed otherwise. 


All medical decisions are between a minimalist-maximalist range, like in quants case it's between nowcasting and forecasting. How to find the balance or at least mostly be on the winning side? (at least on winning side may not be acceptable in medical decision making, as having minimum losses “do no harm” is the first important rule not to be missed when caring for a patient.) 


My understanding is, risk management is the key to finding the right balance which will help minimize losses, in quant field there are various objective tools for that, and same for the AI (and present day HI - human intelligence) is the way to go for medical decision making, so in medicine using techniques to find ADR, Anaphylaxis early and intervene early for them is important, evidence based rational use to prevent overtesting while also avoid undertesting and same way over treatment and under treatment. Also detecting suboptimal response of drugs in individual to find opportunity to replace with better alternatives if possible and even predicting what drugs may response better based on genotype/metabolomics/microbiome more easily and widely and other risk management strategies. These all are already known and application of these must increase towards finding the balance but still all these strategies together also may not guarantee a black swan / disaster in medicine but sometimes for the reward (clinical benefits) we have to take the remaining risk.

An example of black swans in medicine is PSA test misuse and many such examples exist there in medicine, which must be prevented by keeping strong focus on research (data generation , collection and analysis) for evidence and its appraisal.

The black swan / disaster in medicine may not be a sudden bad event, rather a bad practice not caught and corrected early.

Read full series here -
Trading and Medicine analogy (complexity, uncertainty, technology) - - https://classworkdecjan.blogspot.com/2021/07/trading-and-medicine-analogy-complexity_52.html

No comments:

Post a Comment