AdSense

Wednesday, June 8, 2016

The Black Swan / Nassim Nicholas Taleb

Features of the "black swan":
1. Abnormal (outside of our imagination)

2. Huge impact

3. Two types.
Type I: After it happened, we tend to think it predictable by using reasons that sound like a truth (but actually aren't). Probabilties / impacts of this type are overestimated.
Type II: We cannot include this type into our model, so it tends to be ignored. Probabilties / impacts of this type are underestimated.


We have to distinguish two things:
[A] There is no evidence that a "black swan" exists.
[B] There is an evidence that a "black swan" does not exist.


You can keep betting on a black swan, if you want. However, people tend to prefer small and short-term continuous profits to big ones that might not get realized in the long run.


Consider biases (e.g., survivorship bias - You can't know what underdogs did/thought.) and distortion (e.g., skewness, kurtosis, over- and under-estimation, estimates vs realities).


You must imagine what you cannot see or recall. People do not recognize what they cannot see and something don't promote their emotional interests.


Do right things at a rough estimate and/or hypothesis, don't do wrong things in a precise manner.

The Black Swan: Second Edition: The Impact of the Highly Improbable Fragility" (Incerto) Kindle Edition by Nassim Nicholas Taleb
amazon.com
amazon.co.jp

No comments:

Post a Comment

Deep Learning (Regression, Multiple Features/Explanatory Variables, Supervised Learning): Impelementation and Showing Biases and Weights

Deep Learning (Regression, Multiple Features/Explanatory Variables, Supervised Learning): Impelementation and Showing Biases and Weights ...