Bayesian Reasoning
692 views | +0 today
Follow
Your new post is loading...
Your new post is loading...
Scooped by Marco Valli
Scoop.it!

Portfolio Management under Stress: A Bayesian-Net Approach to Coherent Asset Allocation: Riccardo Rebonato, Alexander Denev: 9781107048119: Amazon.com: Books

Portfolio Management under Stress: A Bayesian-Net Approach to Coherent Asset Allocation

~ Alexander Denev (author) More about this product
List Price: $65.00
Price: $58.50
You Save: $6.50 (10%)
Portfolio Management under Stress: A Bayesian-Net Approach to Coherent Asset Allocation [Riccardo Rebonato, Alexander Denev] on Amazon.com. *FREE* shipping on qualifying offers. Portfolio Management under Stress offers a novel way to apply the well-established Bayesian-net methodology to the important problem of asset allocation under conditions of market distress or
Marco Valli's insight:

Disclaimer: I personally know one of the authors (AD), who happens to be a friend of mine, as we both took our Master Degree in Physics at "Sapienza", University of Rome. He's a really smart guy and we discussed a little bit, in recent times, about Bayesian reasoning: he definitely got the point! Unfortunately, I'm not in the position of saying anything about the "finance" stuff the book is about! You will have to read it yourselves...and then tell me if it's worthy!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Simpson's Paradox

Simpson's Paradox | Bayesian Reasoning | Scoop.it

Simpson's paradox explained, from the original Berkeley University sex-discrimination blunder.

Marco Valli's insight:

In each of seven University Department, men seems to overperform women in applications. However, aggregate results clearly indicate that women are doing better than males! How is it possible? Simposon's  Paradox is at work! This is a nice graphic explanation of the correct reasoning (figures are better than the text).

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

An example of why data will not end debate - Big Data, Plainly Spoken (aka Numbers Rule Your World)

An example of why data will not end debate - Big Data, Plainly Spoken (aka Numbers Rule Your World) | Bayesian Reasoning | Scoop.it
One oft-repeated "self-evident" tenet of Big Data is that data end all debate. Except if you have ever worked for a real company (excluding those ruled by autocrats), and put data on the table, you know that the data do not end anything. Reader Ben M. sent me to this blog post by Benedict Evans, showing a confusing chart showing how Apple has "passed" Microsoft. Evans used to be a stock analyst before moving to Andreessen Horowitz, a VC (venture capital) business. He has over 25,000 followers on Twitter. I'll get to this chart later but feel free to tour...
Marco Valli's insight:

Quoting from the first lines: "One oft-repeated self-evident tenet of Big Data is that data end all debate. [...] if you have ever worked for a real company [...] and put data on the table, you know that the data do not end anything [...]". I do completely agree! Data are good, big data are terrific. But  you still need the ability to analyze and model data to extract some value and a story!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Scientific method: Statistical errors

Scientific method: Statistical errors | Bayesian Reasoning | Scoop.it
P values, the 'gold standard' of statistical validity, are not as reliable as many scientists assume.
Marco Valli's insight:

Glad to see an article against the blind use of p-values published in Nature! So sad to think that way too many scientists and researchers just play it all around until the get - by mere chance - a "significant" result... 

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

How to Remember the Key Bayes Formula in Statistics

How to Remember the Key Bayes Formula in Statistics | Bayesian Reasoning | Scoop.it
Mathematics is wonderful! (RT @datameer: How to Remember the Key Bayes Formula in Statistics http://t.co/n15s0CjBs9)
Marco Valli's insight:

This is a longer introduction to Bayesian reasoning, highlighting key ideas and the main differences with the frequentistic approach: in the latter, some of the common questions we would like to answer using probabilistic reasoning are simply ill-posed, as in frequentistic view probability can only be related to repeatable events.

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Big data misused to justify vaccination

Big data misused to justify vaccination | Bayesian Reasoning | Scoop.it
The following article published in Forbes on January 23, Big Data Crushes Anti-Vaccination Movement, illustrates why data should be processed and interpreted b…
Marco Valli's insight:

I don't want (really don't want!) to dwell on the conclusions of the article, but the first two paragraphs highlight a very serious issue. Not every dataset is `big data', and data (whether big or not) can't tell you any story at all. It takes someone skilled and used to model, interpret and analyze them to draw conclusions of any sort! Let Data Scientist come in!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Probability of an Up Year When January is Up | Price Action Lab Blog

Probability of an Up Year When January is Up | Price Action Lab Blog | Bayesian Reasoning | Scoop.it
Marco Valli's insight:

In the first half, this article clearly states the limitations and naive-ty of a simple application of Bayes rule to the stock market, stressing more than one time the big assumptions involved. Nevertheless, I'm strongly disagree with statements such as "probability are of little use, unless the probability is 1 or 0", or "probability and risk are loosely related". Truth is, probability and "risk" are not, on their own, a firm basis to make decision. A complete bayesian analysis and utility consideration are due to a complete assessment of the problem, to support a fully-informed decision making process.

more...
No comment yet.
Rescooped by Marco Valli from Statistics- Bayes or Frequentist? Induct, Deduct or Abduct?
Scoop.it!

Icons for the essence of Bayesian and frequentist data analysis

Icons for the essence of Bayesian and frequentist data analysis | Bayesian Reasoning | Scoop.it
... x-y scatter plots as diagrams of data, and many people are familiar with lines as a model form. The credible parameters can then be suggested by a smattering of lines sampled from the posterior distribution, like this (Fig.

Via RMohan
Marco Valli's insight:

As always, an instructive insight from John Kruschke's "Doing Bayesian Data Analysis", that clearly demonstrates in a visual fashion the key difference between the two approaches to uncertainty. Once more, the Bayesian method emerges as the one much closer to what our brain "instinctively" does...

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Monte Carlo Methods in Climate Science

http://www.maa.org/

Marco Valli's insight:

Application of Bayesian reasoning to climate science. It's not my field, so I am asking you whether if fits or not...

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

The Squirrel's Nest: Understanding Bayes, and Economics

The Squirrel's Nest: Understanding Bayes, and Economics | Bayesian Reasoning | Scoop.it
In a recent NY Times piece the Fellow from the South had a discussion on Bayes analysis. Now I wrote a book on that in the late 1960s so let me refresh your memory. Bayes said that if you took into account the data up to the ...
Marco Valli's insight:

Efficient Market Hypothesis and related topics...

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Probably Overthinking It: Leslie Valiant is probably British. Or old.

Probably Overthinking It: Leslie Valiant is probably British. Or old. | Bayesian Reasoning | Scoop.it
Marco Valli's insight:

A nice application of Bayesian reasoning: that's just the way our brain works!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

A quotable Domingos paper

A quotable Domingos paper | Bayesian Reasoning | Scoop.it
I've been (re-)reading Pedro Domingos' paper, A Few Useful Things to Know About Machine Learning, and wanted to share some quotes that I like. (...) much of the "folk knowledge" that is needed to s...
Marco Valli's insight:

There nothing *strictly* bayesian here, but a lot of wisdom, indeed...the lacking of "folk knowledge" in textbooks is true for Bayesian Networks, too...and the "incorporating knowldegde" tale is a key feature of the bayesian approach!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Really Bayesian logistic regression in Python | This Number Crunching Life

Really Bayesian logistic regression in Python | This Number Crunching Life | Bayesian Reasoning | Scoop.it
Marco Valli's insight:

"[...] By really Bayesian, I mean that I want a distribution over model parameters as my answer [...]". So be it!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Nate Silver on Bayes' Theorem and the power of big data done right

Nate Silver on Bayes' Theorem and the power of big data done right | Bayesian Reasoning | Scoop.it
Nate Silver outlines three principles taken from Bayes' Theorem that influence his prediction process at the Gartner BI Summit in Dallas, Texas.
Marco Valli's insight:

Nate Silver star has been on the rise since 2012 presidential election prediction...but how many of those who cites him clearly understand what he has done and how the Bayesian approach is related to this story?!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Conditional probability

Conditional probability | Bayesian Reasoning | Scoop.it

A visual explanation of conditional probability. From setosa.io

Marco Valli's insight:

A nice visualization of Bayes theorem and conditional probabilities. Fully playable. Have fun!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Doing Bayesian Data Analysis: Improved icons for Bayesian and frequentist analysis

Doing Bayesian Data Analysis: Improved icons for Bayesian and frequentist analysis | Bayesian Reasoning | Scoop.it
Marco Valli's insight:

Yet another clear and visual display of the differences between the frequentist and bayesian approach to uncertainty, hypothesis testing and parameter estimation. And the winner is...

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Disinformation visualization

Disinformation visualization | Bayesian Reasoning | Scoop.it

Flowing Data

Marco Valli's insight:

<<It comes back to the whole "let the data speak" ideal. Data might have something to say, but the analyst, designer, etc still has to translate, whether that's through statistical methods or visualization. Sometimes meaning gets lost when you're not careful.>> No need for any insight from my side: that's the whole story! Be careful! Distrust the one who says "let the data speak for themselves"!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

An Intuitive (and Short) Explanation of Bayes’ Theorem | BetterExplained

Bayes' theorem was the subject of "a detailed article":http://www.yudkowsky.net/bayes/bayes.html. The essay is good, but over 15,000 words long -- here's the (one more explanation of Bayes Theorem for #mgmt228 http://t.co/3CflTyw5UI)...
Marco Valli's insight:

A (not so) short intro; things to keep in mind when trying to explain this topic!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Microsoft’s Duncan Watts on Decisions, Data, and the Scientific Method | MIT Technology Review

Microsoft’s Duncan Watts on Decisions, Data, and the Scientific Method | MIT Technology Review | Bayesian Reasoning | Scoop.it
More than technology, businesses need the scientific method.
Marco Valli's insight:

This is a key point, that the "big data" era is largely overlooking. Data, even if "big", can't tell the whole story. That's when the "data scientist" comes in! Great to have big data, but what you definitely need is someone who knows what to do with them, how to build up a model and how to interpret outcomes. Bayesian reasoning perfectly fits to those needs.

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Probably Overthinking It: The Red Line problem

Probably Overthinking It: The Red Line problem | Bayesian Reasoning | Scoop.it
Marco Valli's insight:

A simple application of Bayesian reasoning to a common, every-day-life problem: undergroung waiting times! Key point is the subtle difference with "perceived" and "actual" waiting times (the same applies to overcrowded restaurants, buses, flights...)! Enjoy 

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

The Potential of Social Network Analysis in Intelligence

The Potential of Social Network Analysis in Intelligence | Bayesian Reasoning | Scoop.it
Within its limits, SNA can be applied to identify individuals or organizations within a network, generate new leads and simulate flows of information or money throughout a network.
Marco Valli's insight:

SNA and NSA...

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

Jamesian Philosophy Refreshed: A few words about Thomas Bayes

Jamesian Philosophy Refreshed: A few words about Thomas Bayes | Bayesian Reasoning | Scoop.it
Thomas Bayes, a mid-eighteenth century English mathematician, dead since 1761, performed magic the ramifications of which continue to unfold, and in directions of interest to the world of finance today.
Marco Valli's insight:

A very short, but clear presentation of Bayes Theorem...and why it should be more properly called Laplace Theorem; a funny application to poker, too...

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

The Inference Button: Bayesian GLMs made easy with PyMC3

Marco Valli's insight:

Python tool for bayesian linear regression. I appreciate the clear stating of two major advantages of bayesian reasoning in parameter estimation, namely 1) quantified uncertainty and 2) priors. Nevertheless, the first issue is really and advantage. The "priors" stuff is a "need". I can't believed even trained physicist still just put down models without constraining parameter space by using what they DO know...

more...
No comment yet.
Rescooped by Marco Valli from Statistics- Bayes or Frequentist? Induct, Deduct or Abduct?
Scoop.it!

Top 20 things politicians need to know about science - The Guardian

Top 20 things politicians need to know about science - The Guardian | Bayesian Reasoning | Scoop.it
The Guardian
Top 20 things politicians need to know about science
The Guardian
Thus P = 0.01 means there is a 1-in-100 probability that what looks like an effect of the treatment could have occurred randomly, and in truth there was no effect at all.

Via RMohan
Marco Valli's insight:

...not just politicians! I'd put some of the issues in a slightly different lights, but the key points are all there!

more...
No comment yet.
Scooped by Marco Valli
Scoop.it!

DATING FOR BAYESIANS: Here's How To Use Statistics To Improve Your Love Life

DATING FOR BAYESIANS: Here's How To Use Statistics To Improve Your Love Life | Bayesian Reasoning | Scoop.it
I am a somewhat socially awkward person. This...
Marco Valli's insight:

A naive but not oversimplified introduction to the subject! Enjoy your dates!

more...
No comment yet.