TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

What Is Bayesian/Frequentist Inference? (2012)

74 pointsby spekcularover 2 years ago

8 comments

clircleover 2 years ago
This blog is by Larry Wasserman, so i think his advice should be taken seriously. I agree that there are uses of both philosophies, and that statisticians should be pragmatic rather than dogmatic.<p>My issue is that his advice is most useful for statisticians working in the abstract, but it doesn’t really help people working with real data. Scientists and data analysts just want to know how to analyze their data, and this does not help them. I know that stats isn’t a cookbook, but we could put some guard rails down that help practitioners with their problems.
评论 #33154844 未加载
评论 #33155599 未加载
评论 #33159561 未加载
wirrbelover 2 years ago
Dunno whether I agree to this. I agree that both are acceptable ways to do statistics. However<p>1. Bayesian stats is an approach that tends to make model assumptions fairly explicity, whereas in frequentist approaches, many assumptions are fairly implicit (Normal distribution of data, etc.) 2. I would consider myself a Bayesianist but I am sceptical about too much mention of esoteric terminology like &quot;Belief&quot;. Bayesian probabilities are probabilities following the Kolmogorov axioms, which is also the foundation of Frequentist stats.<p>For decades, Bayesian inference was impractical because we need to resort to sampling methods and (a) computational power was insuffient and (b) we didn&#x27;t have algorithms like No U-Turn Sampler (NUTS).<p>Both aspects are &#x27;solved&#x27;, so why is Bayesianism not universally adopted? Of course it still has a reputational disadvantage, but I think more importantly its because<p>* frequentist methods are good enough for purposes of publishing research [<i>] </i> for some problems we really have a hard time assembling bayesian graphs * some inference methods (e..g. Kalman filter) can both be seen as frequentist or Bayesian<p>As a bayesianist I am amazed at how well frequentism can work, even when the &#x27;traditional&#x27; way of applying it contradicts the derivations of founding fathers like Fisher&#x2F;Pearson. It&#x27;s almost as if we have an evolutionary process at play.<p>[*] That is, if you use p-Values as publication thresholds
评论 #33156202 未加载
tpoacherover 2 years ago
Fully agree with every single word in this article. Particularly the bit about &quot;identity statistics&quot;.<p>Also, regarding failure of notation: I&#x27;ve been arguing for a while that the notation we use for probability is highly problematic and effectively eggregious abuse of notation. And not just the whole &quot;belief vs frequency of hypotheticals&quot;, but even the simple fact of what it represents. Does p(x) denote an event? The whole distribution? The distribution over a particular space?<p>An apologist might rightly point out that p(x) is actually shorthand for p(x=X), where x is the event and X is the distribution. But even this screams confusion.<p>For me the ideal notation would have been something that makes it explicit that the probability describes a set of &#x27;sampling events&#x27; from a &#x27;pool&#x27;, e.g. Prob_{x}(X), where x is the set of events, and X is the nature of the distribution (i.e. a function which returns frequencies&#x2F;beliefs over a domain).<p>And probability <i>density</i> functions would be denoted as <i>actual</i> derivatives, e.g. d&#x2F;dx P_x (Cum.Normal)
enaaemover 2 years ago
The difference between Bayesian and Frequentist is in the interpretation of randomness. In Bayesian statistics &#x27;randomness&#x27; is not a property of nature but a description of our knowledge.<p>What&#x27;s randomness in a coin toss? If we had all the information we could perfectly predict the result of a toss. But if we know nothing then at most we can say is that both outcomes are equally probable.<p>Another example, if you had no idea who the next presidential winner will be between two candidates, than saying it&#x27;s 50-50 is an accurate description of your knowledge.<p>If anyone is more interested I would refer to you to [1]. Here, probability theory is interpreted as an extension of logic. Very interesting stuff.<p>[1] <a href="http:&#x2F;&#x2F;www.med.mcgill.ca&#x2F;epidemiology&#x2F;hanley&#x2F;bios601&#x2F;GaussianModel&#x2F;JaynesProbabilityTheory.pdf" rel="nofollow">http:&#x2F;&#x2F;www.med.mcgill.ca&#x2F;epidemiology&#x2F;hanley&#x2F;bios601&#x2F;Gaussia...</a>
评论 #33240967 未加载
评论 #33157304 未加载
评论 #33157232 未加载
lalaithionover 2 years ago
This falls apart in higher dimensions, but in the example given in the article the two answers only differ because they have different priors. If you repeat the bayesian analysis using the prior \theta ~ N(0, x), and let x go to infinity, then you approach the frequentist answer.<p>In my opinion, <a href="https:&#x2F;&#x2F;stats.stackexchange.com&#x2F;questions&#x2F;2272&#x2F;whats-the-difference-between-a-confidence-interval-and-a-credible-interval" rel="nofollow">https:&#x2F;&#x2F;stats.stackexchange.com&#x2F;questions&#x2F;2272&#x2F;whats-the-dif...</a> is a better explanation of the difference between confidence and credible intervals.<p>Editing to add more commentary to my link:<p>If you&#x27;ve read the link, one of the principal objections of the frequentist is &quot;What if the jar is type B? Then your interval will be wrong 80% of the time, and only correct 20% of the time!&quot;<p>This is because, if you look at the original numbers, jar B has all types of cookies, and therefore any single draw from jar B can &quot;look like&quot; any other jar, and because other jars have more concentrated cookie types, they are &quot;more likely&quot; answers for each potential sample.<p>This issue also comes up with the frequentist analysis! If you look at the confidence intervals, they all say &quot;This jar could be jar B&quot;. Instead of being bad at detecting jar B, they are good at always considering jar B no matter the evidence – but it&#x27;s the same uncertainty.<p>The bayesian version of this objection is &quot;when you pull a cookie with 3 chips your interval is only correct 41% of the time&quot;. This is because if we get jar A, we&#x27;ll probably draw a 2-chip cookie, so it&#x27;s outside of our confidence interval.<p>But note that we probably don&#x27;t really care about the confidence interval or credibility interval. It&#x27;s basically a hack to take a probabilistic problem and turn our answer into black and white. To say &quot;these hypotheses are valid and these are invalid&quot;.<p>But this is statistics! If you just take the Bayesian approach, and throw out the need to create an arbitrary interval, you can just stop at the table titled P(Jar|Chips). That&#x27;s all the information you need. If you draw a N-chip cookie, you can use that table to update your P(Chips_2) for a second draw, and you&#x27;ll get a concrete probabilistic answer. Yes, you have to assume a prior. But frequentist statistics literally can&#x27;t answer this question! Without a prior, there&#x27;s no way to turn P(Chips | Jar) into a P(Jar | Chips) to update on, so you can&#x27;t track your evidence to get better predictions. You just sit there saying &quot;well, my interval meets the criterion even in the worst case&quot;.
评论 #33169473 未加载
dangover 2 years ago
Discussed at the time:<p><i>What Is Bayesian&#x2F;Frequentist Inference?</i> - <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=4800449" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=4800449</a> - Nov 2012 (27 comments)
sega_saiover 2 years ago
In the end I believe the Bayesian inference is more straightforward to implement and understand if you can afford computationally sampling of the posterior. So I think at least in physics there is a shift towards Bayesian approaches.
dekhnover 2 years ago
Whilst successful in my career and user of probability, statistics, and inference on a regular basis, I simply cannot understand what&#x27;s being discussed here.<p>I don&#x27;t even want to understand it. Just like quantum, half the argument seems to be the a mismatch between mental models and actual reality.
评论 #33155708 未加载