Questions tagged [bayesian]

Bayesian inference is a method of statistical inference that relies on treating the model parameters as random variables and applying Bayes' theorem to deduce subjective probability statements about the parameters or hypotheses, conditional on the observed dataset.

Filter by
Sorted by
Tagged with
0 votes
0 answers
11 views

Why do Bayesians not model the input as a random variable in discrminative models?

From this question: Example of a Discriminative Bayesian Model There is this table comparing Frequentist/Bayesians and Discriminative/Generative models. $$\begin{array}{l|c|c} & \text{Frequentist}...
  • 133
0 votes
0 answers
24 views

The gradient vector in Hamiltonian Monte Carlo (leapfrog method)

Let $x_{t}, \omega_{t} \in \mathbb{R^{d}}$ The Hamiltonian Monte Carlo says this: Deterministic: it relies on the Hamiltonian dynamics so given an initial state, at any time $t$, specified by the ...
  • 207
2 votes
1 answer
21 views

How does maximizing ELBO in Bayesian neural networks give us the correct posterior predictive distribution?

In Bayesian/variational neural networks one often uses the Evidence Lower BOund (ELBO) as the objective function to optimize with respect to the model parameters. That is if $D=\{y_i,x_i\}_{1\dots n}$ ...
0 votes
0 answers
11 views

Conditional Probability expression

I am trying to understand the next expression about a conditional probability: $P(D|T)=P(V|T)P(D|V)+(1-P(V|T))P(D|\neg V)$ It seems a variation of the Bayes theorem, but I can't see how this ...
  • 925
0 votes
0 answers
14 views

Incorporate partial information about Y into predictions

I have a linear regression model predicting exports of toys from the United States on an annual basis. This initial model is based on a few factors: toy companies' demand projections, toy production ...
0 votes
0 answers
22 views

RetroBayesianism or? I have access post-hoc to "surprised\shocked" assessments only. What math to use? [closed]

This is a basic alternative"fringe" statistics question as to who has studied this under what name(s), or what you recommend as a personal\undocumented workaround or adaptation\combination ...
1 vote
0 answers
8 views

Clarification on the connection between deep ensembles and bayesian neural networks

I'm not sure if I understand the relationship between deep ensembles and Bayesian interpretation well. can you tell me if i am right or wrong? Suppose I have an ensemble of L neural networks trained ...
  • 195
0 votes
1 answer
32 views

Interpreting posterior with Half-Normal shape

I am building a Marketing Mix Model in PyMC and am not sure how to interpret the posteriors, especially those with half-normal priors (sigma=1). I’ve chosen this prior because media could not have a ...
0 votes
0 answers
22 views

Why is bambi returning multiple posterior mean predictions per (chain,draw)? [closed]

Let's say I want to calculate posterior for conversion rate per group. y denotes successes, n trials. ...
1 vote
2 answers
39 views

Is it practical to derive the prior distribution by dividing the posterior by the likelihood and multiplying by the "evidence"?

Is it practical to derive the optimal prior distribution by dividing the posterior by the likelihood and multiplying by the "evidence"? Suppose you assume a probability distribution. You ...
0 votes
0 answers
11 views

Reasoning about modelling uncertainty w.r.t input

I am trying to build up my reasoning about uncertainty modelling and ways of modelling it. What I am trying to essentially get at is how changes in input variables results in different posteriors(...
0 votes
0 answers
45 views

Finding the scale of a $t$ distribution given information about CDFs

I'm using this 2017/18 exam paper to prepare for a course in Bayesian Statistics. It has a question, Assume $\{x_1,\dots,x_n\}$ is a random sample from a Gaussian distribution with mean $\mu$ and ...
  • 343
1 vote
0 answers
16 views

How do I multiply two discrete distributions together?

As part of my PhD research, I am simulating a system that comprises multiple sensors, and I want to perform sensor fusion. My sensors give me sets of measurements (values between 0 and 1 mm) that I ...
2 votes
0 answers
82 views

Using the bootstrap estimates to compute the acceleration parameter for BCa bootstrap CI

I have noticed different software use different method of calculating the acceleration parameter, $a$, for computing the bias-corrected and accelerated (BCa) bootstrap confidence interval (CI). In ...
  • 24.3k
0 votes
1 answer
25 views

How to interpret BayesFactor::contingencyTableBF BF results in R?

Say we build the following table: ...
  • 101
0 votes
0 answers
18 views

Using Bayes' Formula, how to obtain class prior and other components from the data

I am trying to solve the following problem using Bayes' formula 'Given visitors have visited a page, what is the probability they will convert?' I have a website with which people interact. It has 50 ...
  • 1
0 votes
0 answers
12 views

Bayesian Model selection vs Model comparison

Is anyone aware of any articles or book chapters about the distinction between model selection and model comparison in bayesian multilevel modeling? I am fitting several competing growth models using ...
0 votes
0 answers
5 views

What is the distinction between models with multivariate priors and (multiplicative) interaction terms?

In some models, groups of parameters are assigned multivariate priors with a covariance matrix where (at least some) elements are estimated. The motivation is that the parameters depend on each other. ...
1 vote
0 answers
16 views

When to specify multivariate versus univariate priors on parameters?

Suppose a linear regression model: $$y \sim Normal(\beta X, \sigma)$$ For our purposes, assume $y$ is a univariate outcome and $X$ is a design matrix containing an intercept and one additional ...
0 votes
0 answers
5 views

Does the following "approach" for integrating over the data space makes sense?

Suppose that we have the following posterior and prior distributions $p(\mu|x,m_{1}(x),s_{1}(x)) = Normal(\mu;m_{1}(x),s_{1}(x))$ and $p(\mu|m_{2},s_{2})$ The $m_{1}(x),s_{1}(x)$ indicate that the ...
  • 1,883
0 votes
0 answers
22 views

Marginal Likelihood Computation for Bayesian Linear Model

Given a simple Bayesian linear model with $N$ observations $y = X\beta + \varepsilon \quad \quad \varepsilon \sim \mathcal{N}(0, \Sigma)$ with known error variance-covariance matrix $\Sigma$ and ...
  • 651
0 votes
0 answers
13 views

Finding posterior with Ga prior, Exp likelihood

I'm using this 2017-2018 exam paper to pre-study for a module in Bayesian Statistics. It has a question, Assume that the waiting time, $t$, of a client in a bank can be modelled with an exponential ...
  • 343
1 vote
1 answer
34 views

In what sense is "Bayesian cross validation" Bayesian?

In cross validation, we repeat training the model with resampled training data and measure average of the errors from the different resampled training samples. So cross validation is basically a ...
1 vote
0 answers
18 views

Connection Between Bayesian Prior and Variable Selection in Lasso [duplicate]

I am interested in learning more about the Bayesian interpretation of the Lasso model. The Lasso model assumes a Laplace distribution of coefficients and the optimal coefficients maximize the ...
  • 213
0 votes
0 answers
34 views

Applying Bayes Theorem to combine probability mass functions of time series changes

I apologize. I'm not well trained in formal statistics, so feel free to gently correct my terminology and methods. For a univariate continuous real-valued time series X, I've calculated probability ...
  • 363
0 votes
0 answers
8 views

Can I calculate precision from prevalence and misclassification rate?

If I know $P(D+)=P(disease+)=0.001$, a test is accurate (meaning the test agrees with the truth) 99% of the times, could I calculate the precision $P(D+|T+)=P(disease+|test+)$? $P(D+|T+)=\frac{P(T+|D+)...
  • 23
2 votes
1 answer
26 views

Finding likelihood given uniform data

I'm using this past paper to pre-study for a module in Bayesian Statistics. It has a question, A precision weighing device yields unbiased measurements within half a gramme, which can be modelled as $...
  • 343
0 votes
2 answers
24 views

Is it possible to use a scaled version of a beta distribution to represent lifetime?

I'm a beginner in statistics, so I'm not sure if this has been asked before. I've looked, but I couldn't find an answer. So I'm trying to use Bayes' theorem to build a probability distribution ...
2 votes
1 answer
23 views

Frequentist method for random samples from unknown urn

Say you have two urns with a large number of red and blue marbles each and you know the proportion of red and blue marbles in each urn. Now we choose one urn at random (but don't know which) and ...
0 votes
0 answers
9 views

How can I conduct bayesian inference analysis from the air pollution dataset? [closed]

I have to come up with the analysis and write bayesian inference python with one of the bayesian topic. I try using this dataset https://www.kaggle.com/datasets/rohanrao/air-quality-data-in-india and ...
2 votes
1 answer
32 views

How to determine interventional distributions from observational data?

How do we compute/query interventional distributions from observational data (i.e. without knowledge of the causal graph such as a Structural Causal Model (SCM))?
1 vote
1 answer
29 views

Difference between sequential and one-batch Bayesian update

I learnt that sequential Bayesian update and one batch (all at once)update will give the same result if the observations are i.i.d. I tried to test this theory using my model which contains 4 ...
  • 13
0 votes
0 answers
12 views

Computing Joint Distribution from Marginal Distributions, and Vice Versa

I'm trying to learn bayesian networks by doing probability computations by hand. Given the probability distribution P(D),P(I),P(G|I,D), I want to compute ...
  • 231
0 votes
0 answers
22 views

Updating prediction by combining confidence values of incoming events

I want to predict a binary variable $y$. I assume a prior probability of $p(y=1) = 0.5$. Now there is evidence from incoming independent “events” that should increase or decrease my overall prediction ...
0 votes
1 answer
50 views

Modelling probabilities of a sum of binomials with different probabilities and trials

I have the following example data, where each row is an independent observation: A B C Y 10 22 6 2 4 60 2 0 12 8 10 3 ... $A$, $B$, $C$ and $Y$ are all positive integers. The variables $A$, $B$ ...
1 vote
1 answer
16 views

How do I include 0-rated items while ranking items with variable number of ratings?

I have a list of items and ratings from 0 to 10, with decimal ratings so that possible ratings are 0, 0.5, 1, 1.5 ... 9.5, 10. I am using https://www.evanmiller.org/ranking-items-with-star-ratings....
0 votes
0 answers
21 views

Bayes rule with two X variables

I'm trying to work out the proper updating of two (independent) random variables ($\pi_a$, $\pi_b$), given an observation $y$. Each $\pi$ is Bernoulli, corresponding to a hidden state. ($\pi_a$, $\...
5 votes
3 answers
797 views

Why not use the same distribution for the prior in Bayesian statistics?

I am wondering why introductory books on statistics use a conjugate distribution family for the prior instead of using the same pdf of the one we are trying to infer the parameters? For example, the ...
  • 487
0 votes
0 answers
14 views

Is my current understanding of the conditional independence assumption in bayesian inference correct?

Standard bayesian inference asssumptions: model $M=\{p(\cdot | \theta) | \theta \in \mathcal{T} \}$ as a prior, choose a distribution $\pi$ on $\mathcal{T}$ then assume observations are generated as ...
4 votes
2 answers
78 views

Can we say the frequentist interpretation of probability is more appropriate in the dice rolling problem?

Suppose we role a dice and see what we get, the sample space is $\{1,2,3,4,5,6\}$ and each outcome occurs with probability 1/6. For example, if we look at the probability that 6 appears, it seems ...
  • 1,898
0 votes
0 answers
14 views

Full conditional distribution (for Gibbs sampling) when model contains a logical node

I'm working on a time series model using a Bayesian implementation in JAGS. The simplest version of my model is an integrated random walk. The heart of the model is this: $y_t \sim N(\mu_t, \sigma)$ ...
4 votes
2 answers
110 views

What is Bayes estimator of $\theta$ when loss function is $L(\theta,a)=I(|\theta -a|>\delta)$?

Suppose $X$ given $\theta$ has pdf $f(x\mid \theta)=e^{-(x-\theta)}I(x>\theta)$ and there is a standard Cauchy prior on $\theta$. As part of an exercise, I am trying to find a Bayes estimator of $\...
  • 9,559
16 votes
4 answers
1k views

Priors that do not become irrelevant with large sample sizes

This may be a weird question. My colleagues and I are working on a medical estimation problem, where relevant prior knowledge regarding plausible values of some physiological parameters exists. In ...
  • 1,484
1 vote
1 answer
65 views

Updated book or review paper on MCMC methods (2022)

For a self-study course, I'm looking for bibliography that describes current MCMC algorithms. I'd prefer a book or a review paper. My background knowledge is at the level of Gamerman-Lopes or Gilks-...
0 votes
0 answers
15 views

Distribution of class sizes above a certain threshold in a stick-breaking process

Given a stick-breaking process, that is to say a sequence of random variables $(Y_n)_n$ such that $$Y_n = X_n \prod_{i=1}^{n-1}(1 - X_i)$$ where the $X_n$ are independent random variables distributed ...
0 votes
0 answers
37 views

Understanding Bayes' Theorem, or: Are all pluviophobes hermits?

Let $A$ be the event "I go out", $B$ be the event "It rains". Then Bayes' Theorem tells us that $$P[A|B] = \frac{P[B|A]P[A]}{P[B]}.$$ I think the weather doesn't care what I do, so ...
  • 343
0 votes
0 answers
8 views

Identification of Latent Variables in Latent Dirichlet Allocation

I am building an Latent Dirichlet Allocation (LDA) model which estimates a version of the Author-Topic model developed by Rosen-Zvi, Griffiths, Steyvers and Smyth. The original paper is here The ...
1 vote
0 answers
13 views

How does one deal with missing data of unbalanced longitudinal data? And what about multivariate case?

For balanced longitudinal data, the missing data can be handled by multiple imputation in wide format where I assume MAR. However, for unbalanced longitudinal data(i.e. number of measurements varying ...
  • 1,159
0 votes
0 answers
8 views

CTR rating of two videos

The studio has launched an advertising campaign for its game. The campaign consisted of two videos, clicking on which the player could go to the game page in the Google Play Store. The results for ...
0 votes
0 answers
11 views

How to select the best binary inputs for a Bayesian model?

I am generating a binary Bayesian model to predict if I am asleep. I set the Prior P(Asleep) based on historical data (8 hours a day = 0.333), and for arguments sake I am using a posterior threshold ...
  • 131

1
2 3 4 5
149