Lab_8_Variational_Inference_PtI.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "authorship_tag": "ABX9TyPGHtq23u2+ORaFVOQr4FWK",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/USCbiostats/PM520/blob/main/Lab_8_Variational_Inference_PtI.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Running Up That Hill, or: Intro to Variational (Bayesian) Inference\n",
        "$\\newcommand{\\data}{\\text{Data}}$\n",
        "$\\newcommand{\\E}{\\mathbb{E}}$\n",
        "Recall that in Bayesian inference, we seek to model the uncertainty in our estimates through a _posterior_ distribution. The posterior is derived from [Bayes' Theorem](https://en.wikipedia.org/wiki/Bayes%27_theorem) as,\n",
        "$$\\Pr(\\theta | \\data) = \\frac{\\Pr(\\data | \\theta) \\Pr(\\theta)}{\\Pr(\\data)},$$\n",
        "where $\\Pr(\\theta | \\data)$ is the [_posterior_ probability](https://en.wikipedia.org/wiki/Posterior_probability) for $\\theta$ and reflects our uncertainty in the values that $\\theta$ may take on, $\\Pr(\\data | \\theta)$ is our likelihood, $\\Pr(\\theta)$ is a [_prior_ probability](https://en.wikipedia.org/wiki/Prior_probability) (or _prior_) over $\\theta$ and $\\Pr(\\data)$ is a [_marginal_ probability/likelihood](https://en.wikipedia.org/wiki/Marginal_likelihood) of the data.\n",
        "\n",
        "Last week, we explored this concept in the regime of \"brute forcing\" the posterior distribution for a simple exercise (e.g., calculating posterior probability an individual is sick, given a positive test) as well as a result in [Exponential Families](https://en.wikipedia.org/wiki/Exponential_family) that leveraged [conjugacy](https://en.wikipedia.org/wiki/Conjugate_prior) (e.g., posterior probability for a coin to land on \"heads\").\n",
        "\n",
        "> *What happens if our model does not have a simple to evaluate or conjugate posterior?*\n",
        "\n",
        "Historically, this is where [Monte-Carlo methods](https://en.wikipedia.org/wiki/Monte_Carlo_method) typically apply, however, these approaches are limited to low- to medium-dimensional applications in practice, and unlikely to scale to ultra high-dimensional settings common in modern datasets (e.g., genetics, genomics, etc)."
      ],
      "metadata": {
        "id": "2Cq4yhMdok8J"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Exact Approximate Inference 🤔\n",
        "\n",
        "Rather than performing inference under an intractible exact posterior $\\Pr(\\theta | \\data)$, we seek to perform inference using a surrogate distribution $Q(\\theta | \\data)$ that is simpler. But how to identify or even quantify how good a proposed surrogate distribution $Q$ is?\n",
        "\n",
        "Recall, [KL-Divergence](https://en.wikipedia.org/wiki/Kullback%E2%80%93Leibler_divergence) captures a notion of \"[statistical distance](https://en.wikipedia.org/wiki/Statistical_distance)\" between parameterized distribution functions, whose definition for discrete variables is,\n",
        "$$\n",
        "\\begin{align*}\n",
        "D_{KL}(q || p) &= \\sum_{x \\in \\mathcal{X}} q(x) \\log \\frac{q(x)}{p(x)} = - \\sum_{x \\in \\mathcal{X}} q(x) \\log \\frac{p(x)}{q(x)}\n",
        "  = -\\mathbb{E}_{x \\sim q}\\left[\\log \\frac{p(x)}{q(x)} \\right].\n",
        "\\end{align*}$$\n",
        "\n",
        "For continuous $x \\in \\mathbb{R}$, we have,\n",
        "$$\\begin{align*}\n",
        "D_{KL}(q || p) &= \\int_{-\\infty}^\\infty q(x) \\log \\frac{q(x)}{p(x)}dx = -\\int_{-\\infty}^\\infty q(x) \\log \\frac{p(x)}{q(x)}dx\n",
        "  = -\\mathbb{E}_{x \\sim q}\\left[\\log \\frac{p(x)}{q(x)} \\right].\n",
        "\\end{align*}$$\n",
        "\n",
        "We can leverage this concept to measure how good a proposal surrogate $Q$ is compared to the true posterior by,\n",
        "$$D_{KL}(Q(\\theta | \\data) || \\Pr(\\theta | \\data))$$\n",
        "however, we often don't know the functional form of $\\Pr(\\theta | \\data)$ let alone compute it in intracable settings! Where are we left?\n",
        "\n",
        "$$\\newcommand{\\ELBO}{\\text{ELBO}}\\begin{align*}\n",
        "D_{KL}(Q(\\theta | \\data) || \\Pr(\\theta | \\data)) &= \\E_Q\\left[ \\log \\frac{Q(\\theta | \\data)}{\\Pr(\\theta | \\data) }\\right] \\\\\n",
        "  &= \\E_Q\\left[ \\log \\frac{Q(\\theta | \\data)\\Pr(\\data)}{\\Pr(\\data | \\theta) \\Pr(\\theta)}\\right] \\\\\n",
        "  &= \\E_Q\\left[ \\log \\frac{Q(\\theta | \\data)}{\\Pr(\\data | \\theta) \\Pr(\\theta)}\\right] + \\E_Q[\\log \\Pr(\\data) ]\\\\\n",
        "  &= \\E_Q\\left[ \\log \\frac{Q(\\theta | \\data)}{\\Pr(\\data | \\theta) \\Pr(\\theta)}\\right] + \\log \\Pr(\\data) \\\\\n",
        "  &= \\underbrace{\\E_Q[ \\log Q(\\theta | \\data)] - \\E_Q[\\log \\Pr(\\data | \\theta)] - \\E_Q[\\log \\Pr(\\theta)]}_{-\\ELBO} + \\log \\Pr(\\data) \\geq 0 ⇒\\\\\n",
        "-\\ELBO \\geq - \\log \\Pr(\\data) \\iff \\ELBO \\leq \\log \\Pr(\\data).\n",
        "\\end{align*}$$\n",
        "\n",
        "The implications of the above derivation suggest that we can maximize the [*evidence lower bound*](https://en.wikipedia.org/wiki/Evidence_lower_bound) or $\\ELBO$ to minimize the KL-divergence between $Q$ and $\\Pr(\\theta | \\data)$ up to a constant.\n",
        "\n",
        "Given this, a helpful representation of the $\\ELBO$ is to re-write it as,\n",
        "$$\\begin{align*}\n",
        "\\ELBO &:= -\\E_Q[ \\log Q(\\theta | \\data)] + \\E_Q[\\log \\Pr(\\data | \\theta)] + \\E_Q[\\log \\Pr(\\theta)] \\\\\n",
        "  &= \\E_Q[\\log \\Pr(\\data | \\theta)] - \\E_Q\\left[ \\log \\frac{Q(\\theta | \\data)}{\\Pr(\\theta)}\\right]\\\\\n",
        "  &= \\E_Q[\\log \\Pr(\\data | \\theta)] + D_{KL}(Q(\\theta | \\data) || \\Pr(\\theta)).\n",
        "\\end{align*}$$\n",
        "This formulation of the $\\ELBO$ can provide helpful intuition as to our objective; which is that we seek to maximize the _expected_ log-likelihood of the data (i.e. $\\E_Q[\\log \\Pr(\\data | \\theta)]$) while penalizing $Q$ that stray too far (in a KL sense) from the prior (i.e. $D_{KL}(Q(\\theta | \\data) || \\Pr(\\theta))$).\n",
        "\n",
        "Fantastic! We've specified how to evaluate our objective (i.e. $\\ELBO$), but we haven't yet specified how to maximize it. As currently stated, we are left with a bit of a puzzle.\n",
        "\n",
        "> *How do we proceed given that $Q$ is a _function_ to identify, and not some fixed variable/parameter?*"
      ],
      "metadata": {
        "id": "kWBxbG1ENsg2"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Variational Inference\n",
        "$\\newcommand{\\indep}{\\perp \\!\\!\\!\\! \\perp}$\n",
        "Before we attempt to maximize the $\\ELBO$, we need to specify the *structural* form of $Q$. Namely, how does $Q$ factorize with respect to our model variables? Typically a *mean-field* factorization is used, which implies that\n",
        "$$Q(\\theta) = \\prod_{j=1}^p Q_j(\\theta_j),$$ or, intuitively that each $\\theta_j \\indep \\theta_{j'}$ for $j \\neq j'$ under $Q$. There are certainly other options for how to factor $Q$ over latent variables (e.g., *structured* mean-field, etc), and trade-offs can sometimes be made over model/computational complexity and downstream accuracy, but often the simplest place to begin is the mean field.\n",
        "\n",
        "Given a factorization for $Q$, we seek to identify the optimal $Q_j^*$ by leveraging optimization techniques from the [calculus of variations](https://en.wikipedia.org/wiki/Calculus_of_variations) (namely the [Euler-Lagrance](https://en.wikipedia.org/wiki/Euler%E2%80%93Lagrange_equation) equation), which tells us that,\n",
        "$$\\begin{align*}\n",
        "\\log Q_j^*(\\theta_j) &= \\E_Q\\left[\\log \\Pr(\\data | \\theta) | \\theta_j\\right] + \\E_Q\\left[\\log \\Pr(\\theta) | \\theta_j \\right].\n",
        "\\end{align*}$$\n",
        "Here, we condition on $\\theta_j$, and compute expectations with respect to $Q$ for _other_ variables $\\theta_{j'}$. Usually, our goal will be to push through some tedious algebra and identify a familiar parametric form that describes $Q_j^*$. However, to do this, we'll need to be able to compute expectations of _other_ approximate posteriors $Q_{j'}^*$, which suggests a cyclic algorithm. In other words, We perform updates for one variable at a time (i.e. [coordinate ascent](https://en.wikipedia.org/wiki/Coordinate_descent) variational inference), or CAVI for short.\n"
      ],
      "metadata": {
        "id": "4c93PQQAOYKv"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Together the CAVI algorithm looks like,\n",
        "\n",
        "![ref Blei D, et al. JASA 2017](data:image/png;base64,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)\n",
        "\n",
        "\n",
        "It is worth noting how different this paradigm is compared to conjugate- or sampling-based Bayesian inference. Here we are performing _optimization_ to identify parametric forms for our surrogate distributions $Q_j$, as opposed to evaluating integrals directly! Hence, we are performing _exact_ optimization for a distribution that _approximates_ the true posterior."
      ],
      "metadata": {
        "id": "aR4TON7Um2BP"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## Example: Bayesian Linear Regression\n",
        "$\\newcommand{\\bX}{\\mathbf{X}}\\newcommand{\\by}{\\mathbf{y}}\\newcommand{\\bI}{\\mathbf{I}}$\n",
        "Let's say we observe $y_i, x_i$ pairs where\n",
        "$$\\begin{align*}\n",
        "y_i | x_i, \\beta &\\sim N(x_i^T \\beta, \\sigma^2) \\\\\n",
        "\\beta_j &\\sim N(0, \\sigma^2_b).\n",
        "\\end{align*}$$\n",
        "\n",
        "If we observe $n$ pairs of $\\{y_i, x_i\\}$, we can re-write the above as\n",
        "$$\\begin{align*}\n",
        "\\by | \\bX, \\beta &\\sim N(\\bX\\beta, \\bI_n \\sigma^2) \\\\\n",
        "\\beta &\\sim N(0, \\bI_p \\sigma^2_b).\n",
        "\\end{align*}$$\n",
        "\n",
        "Here, we'd like to identify an approximate posterior $Q(\\beta)$, rather than its exact posterior $\\Pr(\\beta | \\by)$. Let's take a mean-field approach and assume that $Q$ factorizes as $$Q(\\beta) = \\prod_{j=1}^P Q_j(\\beta_j).$$\n",
        "\n",
        "Next, we need to identify the optimal $Q_j^*$ for each $j$. Namely,\n",
        "$$\\begin{align*}\n",
        "\\log Q_j^*(\\beta_j) &= \\E_Q\\left[\\log \\Pr(\\by | \\beta) | \\beta_j\\right] + \\E_Q\\left[\\log \\Pr(\\beta) | \\beta_j \\right] \\\\\n",
        "  &= \\E_Q\\left[\\log N(\\by | \\bX \\beta, \\bI_n \\sigma^2) | \\beta_j\\right]\n",
        "  + \\E_Q\\left[\\log N(\\beta | 0, \\bI_p \\sigma^2_b) | \\beta_j \\right] \\\\\n",
        "  &= \\E_Q\\left[\\log N(\\by | \\bX \\beta, \\bI_n \\sigma^2) | \\beta_j\\right]\n",
        "  + \\E_Q\\left[-\\frac{1}{2 \\sigma^2_b}\\sum_{j'=1}^P \\beta_{j'}^2 \\ | \\ \\beta_j \\right] \\\\\n",
        "  &= \\E_Q\\left[\\log N(\\by | \\bX \\beta, \\bI_n \\sigma^2) | \\beta_j\\right]\n",
        "  + \\E_Q\\left[-\\frac{1}{2 \\sigma^2_b} \\beta_j^2 \\ | \\ \\beta_j \\right] \\\\\n",
        "  &= \\E_Q\\left[\\log N(\\by | \\bX \\beta, \\bI_n \\sigma^2) | \\beta_j\\right]\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "  &= \\E_Q\\left[-\\frac{1}{2 \\sigma^2}(\\by - \\bX \\beta)^T(\\by - \\bX \\beta) \\ | \\ \\beta_j\\right]\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "\\end{align*}$$\n",
        "$\\newcommand{\\resid}{\\mathbf{r}}$\n",
        "To proceed, let's first define a residual $\\resid_j := \\by - \\sum_{j' \\neq j} \\bX_{j'}\\beta_{j'}$ that reflects residualizing $\\by$ by everything _except_ the $j$th term. Continuing we have,\n",
        "$$\\begin{align*}\n",
        "\\log Q_j^*(\\beta_j) &=\n",
        "  \\E_Q\\left[-\\frac{1}{2 \\sigma^2}(\\by - \\bX \\beta)^T(\\by - \\bX \\beta) \\ | \\ \\beta_j\\right]\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "  &= \\E_Q\\left[-\\frac{1}{2 \\sigma^2}(\\resid_j - \\bX_j \\beta_j)^T(\\resid_j - \\bX_j \\beta_j) \\ | \\ \\beta_j\\right]\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "  &= -\\frac{1}{2 \\sigma^2}\\E_Q\\left[\\resid_j^T\\resid_j - 2\\resid_j^T\\bX_j \\beta_j + \\beta_j^2 X_j^T X_j \\ | \\ \\beta_j\\right]\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "  &= -\\frac{1}{2 \\sigma^2}\\E_Q\\left[ - 2\\resid_j^T\\bX_j \\beta_j + \\beta_j^2 X_j^T X_j \\ | \\ \\beta_j\\right]\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "  &= \\frac{1}{\\sigma^2}\\E_Q[\\resid_j^T]\\bX_j \\beta_j -\\frac{1}{2\\sigma^2}\\beta_j^2 X_j^T X_j\n",
        "  -\\frac{\\beta_j^2}{2 \\sigma^2_b} + O(1) \\\\\n",
        "  &= \\frac{1}{\\sigma^2}\\E_Q[\\resid_j^T]\\bX_j\\beta_j -\\frac{1}{2}\\beta_j^2 \\underbrace{\\left(\\frac{X_j^T X_j}{\\sigma^2}\n",
        "  + \\frac{1}{\\sigma^2_b}\\right)}_{1 / \\widetilde{\\sigma}^2_j} + O(1) \\\\\n",
        "  &= \\frac{1}{\\widetilde{\\sigma}^2_j}\\underbrace{\\frac{\\widetilde{\\sigma}^2_j}{\\sigma^2}\\E_Q[\\resid_j^T]\\bX_j}_{\\tilde{\\mu}_j}\\beta_j -\\frac{\\beta_j^2}{2\\widetilde{\\sigma}^2_j} + O(1) \\\\\n",
        "  &= \\frac{\\tilde{\\mu}_j\\beta_j}{\\widetilde{\\sigma}^2_j} -\\frac{\\beta_j^2}{2\\widetilde{\\sigma}^2_j}\n",
        "  - \\frac{\\tilde{\\mu}_j^2}{2\\widetilde{\\sigma}^2_j} + O(1) \\\\\n",
        "  &= \\frac{(\\beta_j - \\tilde{\\mu}_j)^2}{2 \\widetilde{\\sigma}^2_j} + O(1) \\Rightarrow\\\\\n",
        "Q_j^*(\\beta_j) &= N(\\beta_j | \\tilde{\\mu}_j, \\widetilde{\\sigma}^2_j).\n",
        "\\end{align*}$$\n",
        "This implies that the _optimal_ $Q_j^*(\\beta_j)$ is a normal distribution with variational parameters $\\tilde{\\mu}_j, \\widetilde{\\sigma}^2_j$. Together we have,\n",
        "$$Q(\\beta) = \\prod_j Q_j(\\beta_j) = \\prod_j N(\\beta_j | \\tilde{\\mu}_j, \\widetilde{\\sigma}^2_j).$$\n",
        "\n",
        "Comparing this to the _true_ posterior (which is tractable in this case) we have,\n",
        "$$\\begin{align*}\n",
        "\\Pr(\\beta | \\by) &= N(\\beta | \\widetilde{\\beta}, \\widetilde{\\Sigma}) \\\\\n",
        "\\widetilde{\\beta} &= \\frac{1}{\\sigma^2}\\widetilde{\\Sigma}\\bX^T\\by \\\\\n",
        "\\widetilde{\\Sigma} &= (\\bX^T\\bX\\frac{1}{\\sigma^2} + \\bI_p \\frac{1}{\\sigma^2_b})^{-1}\\\\\n",
        "\\end{align*}$$\n"
      ],
      "metadata": {
        "id": "F1d8YRVNSCYE"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "0J4Da7IKod7S"
      },
      "outputs": [],
      "source": [
        "# Let's code up the CAVI algorithm for bayesian linear regression\n",
        "\n",
        "import jax\n",
        "import jax.numpy as jnp\n",
        "import jax.random as rdm\n",
        "\n",
        "MAX_ITER = 10\n",
        "\n",
        "N = 500\n",
        "P = 250\n",
        "sigma_sq = 0.8\n",
        "sigma_sq_b = 0.1\n",
        "\n",
        "seed = 0\n",
        "key = rdm.PRNGKey(seed)\n",
        "key, x_key, b_key, y_key = rdm.split(key, 4)\n",
        "\n",
        "X = rdm.normal(x_key, shape=(N, P))\n",
        "beta = jnp.sqrt(sigma_sq_b) * rdm.normal(b_key, shape=(P,))\n",
        "y = X @ beta + jnp.sqrt(sigma_sq) * rdm.normal(y_key, shape=(N,))\n",
        "\n",
        "post_means = jnp.zeros((P,))\n",
        "post_vars = jnp.ones((P,)) * sigma_sq_b\n",
        "\n",
        "for _iter in range(MAX_ITER):\n",
        "  ...\n",
        "\n",
        ""
      ]
    }
  ]
}