The document discusses Bayesian data analysis in Python. It begins with an overview of Bayes' theorem and how it relates the prior distribution, likelihood, and posterior distribution. It then provides an example of using a grid approximation to calculate the posterior distribution of a coin's bias given an observed number of heads in tosses. The example generates the prior, likelihood, and normalizes the product of prior and likelihood to get the posterior distribution. It then plots the posterior to show the most likely values of the coin's bias. The document concludes with discussing choosing appropriate prior distributions and conjugate priors that allow direct calculation of the posterior.
The document discusses Bayesian data analysis in Python. It begins with an overview of Bayes' theorem and how it relates the prior distribution, likelihood, and posterior distribution. It then provides an example of using a grid approximation to calculate the posterior distribution of a coin's bias given an observed number of heads in tosses. The example generates the prior, likelihood, and normalizes the product of prior and likelihood to get the posterior distribution. It then plots the posterior to show the most likely values of the coin's bias. The document concludes with discussing choosing appropriate prior distributions and conjugate priors that allow direct calculation of the posterior.
The document discusses Bayesian data analysis in Python. It begins with an overview of Bayes' theorem and how it relates the prior distribution, likelihood, and posterior distribution. It then provides an example of using a grid approximation to calculate the posterior distribution of a coin's bias given an observed number of heads in tosses. The example generates the prior, likelihood, and normalizes the product of prior and likelihood to get the posterior distribution. It then plots the posterior to show the most likely values of the coin's bias. The document concludes with discussing choosing appropriate prior distributions and conjugate priors that allow direct calculation of the posterior.
The document discusses Bayesian data analysis in Python. It begins with an overview of Bayes' theorem and how it relates the prior distribution, likelihood, and posterior distribution. It then provides an example of using a grid approximation to calculate the posterior distribution of a coin's bias given an observed number of heads in tosses. The example generates the prior, likelihood, and normalizes the product of prior and likelihood to get the posterior distribution. It then plots the posterior to show the most likely values of the coin's bias. The document concludes with discussing choosing appropriate prior distributions and conjugate priors that allow direct calculation of the posterior.
Plotting posterior distribution Q: What's the probability of tossing heads with a coin, if we observed 75 heads in 100 tosses? A:
BAYESIAN DATA ANALYSIS IN PYTHON
Let's practice calculating posteriors using grid approximation! B AY E S I A N D ATA A N A LY S I S I N P Y T H O N Prior belief B AY E S I A N D ATA A N A LY S I S I N P Y T H O N
Michal Oleszak Machine Learning Engineer Prior distribution Prior distribution re ects what we know about the parameter before observing any data: nothing → uniform distribution (all values equally likely)
old posterior → can be updated with new data
One can choose any probability distribution as a prior to include external info in the model: expert opinion
common knowledge
previous research
subjective belief
BAYESIAN DATA ANALYSIS IN PYTHON
Prior's impact
BAYESIAN DATA ANALYSIS IN PYTHON
Prior distribution Prior distribution chosen before we see the data.
Prior choice can impact posterior results (especially with li le data).
To avoid cherry-picking, prior choices should be:
clearly stated,
explainable: based on previous research, sensible assumptions, expert opinion, etc.
BAYESIAN DATA ANALYSIS IN PYTHON
Choosing the right prior Our prior belief: heads less likely
Some choices are be er than others!
BAYESIAN DATA ANALYSIS IN PYTHON
Conjugate priors Some priors, multiplied with speci c likelihoods, yield known posteriors.
They are known as conjugate priors.
In the case of coin tossing:
if we choose a prior Beta(a, b),
then the posterior is Beta(#heads + a, #tosses - #heads + b)
Let's practice working with priors! B AY E S I A N D ATA A N A LY S I S I N P Y T H O N Reporting Bayesian results B AY E S I A N D ATA A N A LY S I S I N P Y T H O N
Michal Oleszak Machine Learning Engineer The honest way Report the prior and the posterior of each parameter
