NagoyaStat #12 で使用した資料です(公開に当たって当日ホワイトボードに書いた内容等を補完したものになります)。
「StanとRでベイズ統計モデリング」の第9章前半になります。
第9章のテーマは行列やベクトルを使った演算の高速化です。
---
The title of textbook is "Bayesian statistical modeling with Stan and R", and that of Chapter 9 in textbook is "advanced grammar" in English.
1. The document discusses probabilistic modeling and variational inference. It introduces concepts like Bayes' rule, marginalization, and conditioning.
2. An equation for the evidence lower bound is derived, which decomposes the log likelihood of data into the Kullback-Leibler divergence between an approximate and true posterior plus an expected log likelihood term.
3. Variational autoencoders are discussed, where the approximate posterior is parameterized by a neural network and optimized to maximize the evidence lower bound. Latent variables are modeled as Gaussian distributions.
NagoyaStat #12 で使用した資料です(公開に当たって当日ホワイトボードに書いた内容等を補完したものになります)。
「StanとRでベイズ統計モデリング」の第9章前半になります。
第9章のテーマは行列やベクトルを使った演算の高速化です。
---
The title of textbook is "Bayesian statistical modeling with Stan and R", and that of Chapter 9 in textbook is "advanced grammar" in English.
1. The document discusses probabilistic modeling and variational inference. It introduces concepts like Bayes' rule, marginalization, and conditioning.
2. An equation for the evidence lower bound is derived, which decomposes the log likelihood of data into the Kullback-Leibler divergence between an approximate and true posterior plus an expected log likelihood term.
3. Variational autoencoders are discussed, where the approximate posterior is parameterized by a neural network and optimized to maximize the evidence lower bound. Latent variables are modeled as Gaussian distributions.
AppSocially provides mobile app developers with tools to easily integrate customer acquisition functions into their apps to help grow their user base through viral marketing. Their solution allows developers to make their app inherently viral with just one line of code, track viral metrics and compare to competitors, A/B test "refer friends" campaigns, and hack invitations by including social actions and context to motivate signups, helping developers prioritize marketing without diverting focus from core functions.
Tokyo.R is a community that started in January 2010 that holds meetups to share presentations. It has grown significantly since starting with only 20 subscribers to now having over 200 subscribers. They have had over 726 total presentations presented with about 40% of meetup attendees being new commers and 25% of presentations given by new presenters. Most people only present a few times. The community helps introduce many new people and presenters to the data science field in Tokyo.