The previous post on this blog sought to expose the statistical underpinnings of several machine learning models you know and love. Therein, we made the analogy of a swimming pool: you start on the surface — you know what these models do and how to use them for fun and profit — dive to the bottom — you deconstruct these models into their elementary assumptions and intentions — then finally, work your way back to the surface — reconstructing their functional forms, optimization exigencies and loss functions one step at a time.

In this post, we're going to stay on the surface: instead of deconstructing common models, we're going to further explore the relationships between them — swimming to different corners of the pool itself. Keeping us afloat will be Bayes' theorem — a balanced, dependable yet at times fragile pool ring, so to speak — which we'll take with us wherever we go.

While there are many potential themes of probabilistic models we might explore, we'll herein focus on two: generative vs. discriminative models, and "fully Bayesian" vs. "lowly point estimate" learning. We will stick to the supervised setting as well.

Finally, our pool ring is not a godhead — we are not nautical missionaries brandishing a divine statistical truth, demanding that each model we encounter implement this truth in a rigid, bottom-up fashion. Instead, we'll explore the unique goals, formulations and shortcomings of each, and fall back on Bayes' theorem to bridge the gaps between. Without it, we'd quickly start sinking.

# Discriminative vs. generative models

The goal of a supervised model is to compute the distribution over outcomes $$y$$ given an input $$x$$, written $$P(y\vert x)$$. If $$y$$ is discrete, this distribution is a probability mass function, e.g. a multinomial or binomial distribution. If continuous, it is a probability density function, e.g. a Gaussian distribution.

## Discriminative models

In discriminative models, we immediately direct our focus to this output distribution. Taking an example from the previous post, let's assume a softmax regression which receives some data $$x$$ and predicts a multi-class label red or green or blue. The model's output distribution is therefore multinomial; a multinomial distribution requires as a parameter a vector $$\pi$$ of respective outcome probabilities, e.g. {red: .27, green: .11, blue: .62}. We can compute these individual probabilities via the softmax function, where:

• $$\pi_k = \frac{e^{\eta_k}}{\sum\limits_{k=1}^K e^{\eta_k}}$$
• $$\eta_k = \theta_k^Tx$$
• $$\theta$$ is a matrix of weights which we must infer, and $$x$$ is our input.

### Inference

Typically, we perform inference by taking the maximum likelihood estimate: "which parameters $$\theta$$ most likely gave rise to the observed data pairs $$D = ((x^{(i)}, y^{(i)}), ..., (x^{(m)}, y^{(m)}))$$ via the relationships described above?" We compute this estimate by maximizing the log-likelihood function with respect to $$\theta$$, or equivalently minimizing the negative log-likelihood in identical fashion — the latter better known as a "loss function" in machine learning parlance.

Unfortunately, the maximum likelihood estimate includes no information about the plausibility of the chosen parameter value itself. As such, we often place a prior on our parameter and take the "argmax" over their product. This gives the maximum a posteriori estimate, or MAP.

\begin{align*} \theta_{MAP} &= \underset{\theta}{\arg\max}\ \log \prod\limits_{i=1}^{m} P(y^{(i)}\vert x^{(i)}; \theta)P(\theta)\\ &= \underset{\theta}{\arg\max}\ \sum\limits_{i=1}^{m} \log{P(y^{(i)}\vert x^{(i)}; \theta)} + \log{P(\theta)}\\ \end{align*}

The $$\log{P(\theta)}$$ term can be easily rearranged into what is better known as a regularization term in machine learning, where the type of prior distribution we place on $$\theta$$ gives the type of regularization term.

The argmax finds the point(s) $$\theta$$ at which the given function attains its maximum value. As such, the typical discriminative model — softmax regression, logistic regression, linear regression, etc. — returns a single, lowly point estimate for the parameter in question.

### How do we compute this value?

In the trivial case where $$\theta$$ is 1-dimensional, we can take the derivative of the function in question with respect to $$\theta$$, set it equal to 0, then solve for $$\theta$$. (Additionally, in order to verify that we have indeed obtained a maximum, we should compute a second derivative and assert that its value is negative.)

In the more realistic case where $$\theta$$ is a high-dimensional vector or matrix, we can compute the argmax by way of an optimization routine like stochastic gradient ascent or, as is more common, the argmin by way of stochastic gradient descent.

### What if we're uncertain about our parameter estimates?

Consider the following three scenarios — taken from Daphne Koller's Learning in Probabilistic Graphical Models.

Two teams play 10 times, and the first wins 7 of the 10 matches.

> Infer that the probability of the first team winning is 0.7.

Seems reasonable, right?

A coin is tossed 10 times, and comes out heads on 7 of the 10 tosses.

> Infer that the probability of observing heads is 0.7.

Changing only the analogy, this now seems wholly unreasonable — right?

A coin is tossed 10000 times, and comes out heads on 7000 of the 10000 tosses.

> Infer that the probability of observing heads is 0.7.

Finally, increasing the observed counts, the previous scenario now seems plausible.

I find this a terrific succession of examples with which to convey the notion of uncertainty — that the more data we have, the less uncertain we are about what's really going on. This notion is at the heart of Bayesian statistics and is extremely intuitive to us as humans. Unfortunately, when we compute "lowly point estimates," i.e. the argmin of the loss function with respect to our parameters $$\theta$$, we are discarding this uncertainty entirely. Should our model be fit with $$n$$ observations where $$n$$ is not a large number, our estimate would amount to that of Example #2: a coin is tossed $$n$$ times, and comes out heads on int(.7n) of n tosses — infer that the probability of observing heads is squarely, unflinchingly, 0.7.

### What does including uncertainty look like?

It looks like a distribution — a range of possible values for $$\theta$$. Further, these values are of varying plausibility as dictated by the data we've observed. In Example #2, while we'd still say that $$\Pr(\text{heads}) = .7$$ is the parameter value most likely to have generated our data, we'd additionally maintain that other values in $$(0, 1)$$ are plausible, albeit less so, as well. Again, this logic should be simple to grasp: it comes easy to us as humans.

### Prediction

With the parameter $$\theta$$ in hand prediction is simple: just plug back into our original function $$P(y\vert x)$$. With a point estimate for $$\theta$$, we compute but a single value for $$y$$.

## Generative models

In generative models, we instead compute component parts of the desired output distribution $$P(y\vert x)$$ instead of directly computing $$P(y\vert x)$$ itself. To examine these parts, we'll turn to Bayes' theorem:

$$P(y\vert x) = \frac{P(x\vert y)P(y)}{P(x)}$$

The numerator posits a generative mechanism for the observed data pairs $$D = ((x^{(i)}, y^{(i)}), ..., (x^{(m)}, y^{(m)}))$$ in idiomatic terms; it states that each pair was generated by:

1. Selecting a label $$y^{(i)}$$ from $$P(y)$$. If our model is predicting red or green or blue, $$P(y)$$ is likely a multinomial distribution.
• If our observed label counts are {'red': 20, 'green': 50, 'blue': 30}, we would retrodictively believe this multinomial distribution to have had a parameter vector near $$\pi = [.2, .5, .3]$$.
2. Given a label $$y^{(i)}$$, select a value $$x^{(i)}$$ from $$P(x\vert y)$$. Trivially, this means that we are positing three distinct distributions of this form: $$P(x\vert y=\text{red}), P(x\vert y=\text{green}), P(x\vert y=\text{blue})$$.
• For example, if $$y^{(i)} = \text{red}$$, draw $$x^{(i)}$$ from $$P(x\vert y=\text{red})$$, and so forth.

### Inference

The inference task is to compute $$P(y)$$ and each distinct $$P(x\vert y_k)$$. In a classification setting, the former is likely a multinomial distribution. The latter might be a multinomial distribution or a set of binomial distributions in the case of discrete-feature data, or a set of Gaussian distributions in the case of continuous-feature data. In fact, these distributions can be whatever you'd like, dictated by the idiosyncrasies of the problem at hand.

Finally, we can compute these distributions as per normal: via a maximum likelihood estimate, a MAP estimate, etc.

### Prediction

To compute $$P(y\vert x)$$ we return to Bayes' theorem:

$$P(y\vert x) = \frac{P(x\vert y)P(y)}{P(x)}$$

We have the numerator $$P(y)$$ and three distinct conditional distributions $$P(x\vert y=\text{red}), P(x\vert y=\text{green})$$ and $$P(x\vert y=\text{blue})$$ in hand. What about the denominator?

### Conditional probability and marginalization

The axiom of conditional probability allows us to write $$P(B\vert A)P(A) = P(B, A)$$, i.e. the joint probability of $$B$$ and $$A$$. This is a simple algebraic manipulation. As such, we can rewrite Bayes' theorem in its more compact form.

$$P(y\vert x) = \frac{P(x, y)}{P(x)}$$

Another manipulation of probability distributions is the marginalization operator, which allows us to write:

$$\int P(x, y)dy = P(x)$$

As such, we can marginalize $$y$$ out of the numerator so as to obtain the denominator we require. This denominator is often called the "evidence."

### Marginalization example

Marginalization took me a while to understand. Imagine we have the following joint probability distribution out of which we'd like to marginalize $$A$$.

$$A$$ $$B$$ $$p$$
$$a^1$$ $$b^7$$ $$.03$$
$$a^2$$ $$b^8$$ $$.14$$
$$a^3$$ $$b^7$$ $$.09$$
$$a^1$$ $$b^8$$ $$.34$$
$$a^2$$ $$b^8$$ $$.23$$
$$a^3$$ $$b^8$$ $$.17$$

The result of this marginalization is $$P(B)$$, i.e. "what is the probability of observing each of the distinct values of $$B$$?" In this example there are two — $$b^7$$ and $$b^8$$. To marginalize over $$A$$, we simply:

1. Delete the $$A$$ column.
2. "Collapse" the remaining columns — in this case, $$B$$.

Step 1 gives:

$$B$$ $$p$$
$$b^7$$ $$.03$$
$$b^8$$ $$.14$$
$$b^7$$ $$.09$$
$$b^8$$ $$.34$$
$$b^8$$ $$.23$$
$$b^8$$ $$.17$$

Step 2 gives:

$$B$$ $$p$$
$$b^7$$ $$.03 + .09 = .12$$
$$b^8$$ $$.14 + .34 + .23 + .17 = .88$$

### The denominator

In the context of our generative model with a given input $$x$$, the result of this marginalization is a scalar — not a distribution. To see why, let's construct the joint distribution — the numerator — then marginalize:

$$P(x, y)$$:

$$y$$ $$X$$ $$P(y, X)$$
$$\text{red}$$ $$x$$ $$P(y = \text{red}, x)$$
$$\text{green}$$ $$x$$ $$P(y = \text{green}, x)$$
$$\text{blue}$$ $$x$$ $$P(y = \text{blue}, x)$$

$$\int P(x, y)dy = P(x)$$:

$$X$$ $$P(y, X)$$
$$x$$ $$P(y = \text{red}, x) + P(y = \text{green}, x) + P(y = \text{blue}, x)$$

The resulting probability distribution is over a single value: it is a scalar. This scalar normalizes the respective numerator terms such that:

$$\frac{P(y = \text{red}, x)}{P(x)} + \frac{P(y = \text{green}, x)}{P(x)} + \frac{P(y = \text{blue}, x)}{P(x)} = 1$$

This gives $$P(y\vert x)$$: a valid probability distribution over the class labels $$y$$.

### Partition function

$$P(x)$$ often takes another name and even another variable: $$Z$$, the partition function. The stated purpose of this function is to normalize the numerator such that the above summation-to-1 holds. This normalization is necessary because the numerators typically will not sum to 1 themselves, which follows logically from the fact that:

\begin{align*} \sum\limits_{k = 1}^K P(y = k) = 1 \end{align*}
\begin{align*} P(x\vert y = k) \neq 1 \end{align*}

Since $$(1)$$ is always true, the "$$\neq$$" in $$(2)$$ would need to become an "$$=$$" such that:

$$\sum\limits_{k = 1}^K P(y = k)P(x\vert y = k) = 1$$

Unfortunately, $$P(x\vert y = k) = 1$$ is rarely if ever the case.

As you'll now note, the $$x$$-specific partition function gives a result equivalent to that of the marginalized-over-$$y$$ joint distribution: a scalar value $$P(x)$$ with which to normalize the numerator. However, crucially, please keep in mind:

• The partition function is a specific component of a probabilistic model. It always yields a scalar.
• Marginalization is a much more general operation performed on a probability distribution, which yields a scalar only when the remaining variable(s) are homogeneous, i.e. each remaining column contains a single distinct value.
• In the majority of cases, marginalization will simply yield a reduced probability distribution over many value configurations, similar to the $$P(B)$$ example above.

### In practice, this is superfluous

If we neglect to compute $$P(x)$$, i.e. if we don't normalize our joint distributions $$P(x, y = k)$$, we'll be left with an invalid probability distribution $$\tilde{P}(y\vert x)$$ whose values do not sum to 1. This distribution might look like P(y|x) = {'red': .00047, 'green': .0011, 'blue': .0000853}. If our goal is to simply compute the most likely label, taking the argmax of this unnormalized distribution works just fine. This follows trivially from our Bayesian pool ring:

$$\underset{y}{\arg\max}\ \frac{P(x, y)}{P(x)} = \underset{y}{\arg\max}\ P(x, y)$$

# "Fully Bayesian learning"

We previously lamented the shortcomings of "lowly point estimates" and sang the praises of inferring the full distribution instead. Unfortunately, this is often a computationally-hard thing to do.

To see why, let's revisit Bayes' theorem. Assume we are estimating the parameters $$\theta$$ of a softmax regression model and have placed a prior on $$\theta$$. In concrete terms, this estimate can be written as $$P(\theta\vert D = ((x^{(i)}, y^{(i)}), ..., (x^{(m)}, y^{(m)})))$$: the distribution over our belief in the true value of $$\theta$$ given the data we've observed. Bayes' theorem allows us to expand this quantity into:

$$P(\theta\vert D) = \frac{P(D\vert\theta)P(\theta)}{P(D)}$$

Previously, we computed a "lowly point estimate" for this distribution — the MAP — as:

\begin{align*} \theta_{MAP} &= \underset{\theta}{\arg\max}\ \log \prod\limits_{i=1}^{m} P(y^{(i)}\vert x^{(i)}; \theta)P(\theta)\\ &= \underset{\theta}{\arg\max}\ \log \prod\limits_{i=1}^{m} P(\theta\vert (y^{(i)}, x^{(i)}))\\ \end{align*}

While $$P(y^{(i)}\vert x^{(i)}; \theta)P(\theta) \neq P(\theta\vert (y^{(i)}, x^{(i)}))$$, the argmaxes of the respective products are equal. For this reason, we were able to compute a point estimate for $$P(\theta\vert D)$$, i.e. a "summarization" of $$P(\theta\vert D)$$ in a single value, without ever computing the denominator $$P(D)$$.

(As a brief aside, please note that we could summarize $$P(\theta\vert D)$$ with any single value from this distribution. We often select the maximum likelihood estimate — the single value of $$\theta$$ that most likely gave rise to our data, or the MAP — the single value of $$\theta$$ that both most likely gave rise to our data and most plausibly occurred itself.)

To compute $$P(\theta\vert D)$$ — trivially, a full distribution as the term suggests — we will need to compute $$P(D)$$ after all. As before, this can be accomplished via marginalization:

\begin{align*} P(\theta\vert D) &= \frac{P(D\vert\theta)P(\theta)}{P(D)}\\ &= \frac{P(D, \theta)}{P(D)}\\ &= \frac{P(D, \theta)}{\int P(D, \theta)d\theta}\\ \end{align*}

Since $$\theta$$ takes continuous values, we can no longer employ the "delete and collapse" method of marginalization in discrete distributions. Furthermore, in all but trivial cases, $$\theta$$ is a high-dimensional vector or matrix, leaving us to compute a "high-dimensional integral that lacks an analytic (closed-form) solution — the central computational challenge in inference."1

As such, computing the full distribution $$P(\theta\vert D)$$ is approximating the full distribution $$P(\theta\vert D)$$. To this end, we'll introduce two new families of algorithms.

## Markov chain monte carlo

In small to medium-sized models, we often take an alternative ideological approach to approximating $$P(\theta\vert D)$$: instead of computing a distribution, i.e. the canonical parameters of a gory algebraic expression which control its shape — we produce samples from this distribution. Roughly speaking, the aggregate of these samples then gives, retrodictively, the distribution itself. The general family of these methods is known as Markov chain monte carlo, or MCMC.

In simple terms, MCMC inference for a given parameter $$\phi$$ works as follows:

1. Initialize $$\phi$$ to some value $$\phi_{\text{current}}$$.
2. Compute the prior probability of $$\phi_{\text{current}}$$ and the probability of having observed our data under $$\phi_{\text{current}}$$$$P(\phi_{\text{current}})$$ and $$P(D\vert \phi_{\text{current}})$$, respectively. Their product gives $$P(D, \phi_{\text{current}})$$ — the joint probability of having observed the proposed parameter value and our observed data given this value.
3. Add $$\phi_{\text{current}}$$ to a big green plastic bucket of "accepted values."
4. Propose moving to a new, nearby value $$\phi_{\text{proposal}}$$. This value is drawn from an entirely separate sampling distribution which bears no influence on our prior $$P(\phi)$$ nor likelihood function $$P(D\vert \phi)$$. Repeat Step 2 using $$\phi_{\text{proposal}}$$ instead of $$\phi_{\text{current}}$$.
5. Walk the following tree:
• If $$P(D, \phi_{\text{proposal}}) \gt P(D, \phi_{\text{current}})$$:
• Set $$\phi_{\text{current}} = \phi_{\text{proposal}}$$.
• Move to Step 3.
• Else:
• With some small probability:
• Set $$\phi_{\text{current}} = \phi_{\text{proposal}}$$.
• Move to Step 3.
• Else:
• Move to Step 4.

After collecting a few thousand samples — and discarding the first few hundred, in which we drunkenly amble towards the region of high joint probability (a quantity proportional to the posterior probability) — we now have a bucket of samples from our desired posterior distribution. Nota bene: we never had to touch the high-dimensional integral $$\int P(D, \theta)d\theta$$.

## Variational inference

In large-scale models, MCMC methods are often too slow. Conversely, variational inference provides a framework for casting the problem of posterior approximation as one of optimization — far faster than a sampling-based approach. This yields an analytical approximation to $$P(\theta\vert D)$$. The following explanation of variational inference is taken largely from a previous post of mine: Transfer Learning for Flight Delay Prediction.

For our approximating distribution we'll choose one that is simple, parametric and familiar: the normal (Gaussian) distribution, parameterized by some set of parameters $$\lambda$$.

$$q_{\lambda}(\theta\vert D)$$

Our goal is to force this distribution to closely resemble the original; the KL divergence quantifies their difference:

$$KL(q_{\lambda}(\theta\vert D)\Vert P(\theta\vert D)) = \int{q_{\lambda}(\theta\vert D)\log\frac{q_{\lambda}(\theta\vert D)}{P(\theta\vert D)}d\theta}$$

To this end, we compute its argmin with respect to $$\lambda$$:

$$q_{\lambda}^{*}(\theta\vert D) = \underset{\lambda}{\arg\min}\ \text{KL}(q_{\lambda}(\theta\vert D)\Vert P(\theta\vert D))$$

Expanding the divergence, we obtain:

\begin{align*} KL(q_{\lambda}(\theta\vert D)\Vert P(\theta\vert D)) &= \int{q_{\lambda}(\theta\vert D)\log\frac{q_{\lambda}(\theta\vert D)}{P(\theta\vert D)}d\theta}\\ &= \int{q_{\lambda}(\theta\vert D)\log\frac{q_{\lambda}(\theta\vert D)P(D)}{P(\theta, D)}d\theta}\\ &= \int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D) -\log{P(\theta, D)} + \log{P(D)}}\bigg)d\theta}\\ &= \int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D)} -\log{P(\theta, D)}}\bigg)d\theta + \log{P(D)}\int{q_{\lambda}(\theta\vert D)d\theta}\\ &= \int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D)} -\log{P(\theta, D)}}\bigg)d\theta + \log{P(D)} \cdot 1 \end{align*}

Since only the integral depends on $$\lambda$$, minimizing the entire expression with respect to $$\lambda$$ amounts to minimizing this term. Incidentally, the opposite (negative) of this term is called the ELBO, or the "evidence lower bound."

$$ELBO(\lambda) = -\int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D)} -\log{P(\theta, D)}}\bigg)d\theta$$

To see why, let's plug the ELBO into the equation above and solve for $$\log{P(D)}$$:

$$\log{P(D)} = ELBO(\lambda) + KL(q_{\lambda}(\theta\vert D)\Vert P(\theta\vert D))$$

In English: "the log of the evidence is at least the lower bound of the evidence plus the divergence from our (variational) approximation of the posterior $$q_{\lambda}(\theta\vert D)$$ to our true posterior $$P(\theta\vert D)$$."

As such, minimizing this divergence is equivalent to maximizing the ELBO, as:

$$KL(q_{\lambda}(\theta\vert D)\Vert P(\theta\vert D)) = -ELBO(\lambda) + \log{P(D)}$$

### Optimization

Let's restate the equation for the ELBO and rearrange further:

\begin{align*} ELBO(\lambda) &= -\int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D)} -\log{P(\theta, D)}}\bigg)d\theta\\ &= -\int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D)} -\log{P(D\vert \theta)} - \log{P(\theta)}}\bigg)d\theta\\ &= -\int{q_{\lambda}(\theta\vert D)\bigg(\log{q_{\lambda}(\theta\vert D)} - \log{P(\theta)}}\bigg)d\theta + \log{P(D\vert \theta)}\int{q_{\lambda}(\theta\vert D)d\theta}\\ &= -\int{q_{\lambda}(\theta\vert D)\log{\frac{q_{\lambda}(\theta\vert D)}{P(\theta)}}d\theta} + \log{P(D\vert \theta)} \cdot 1\\ &= \log{P(D\vert \theta)} -KL(q_{\lambda}(\theta\vert D)\Vert P(\theta))\\ \end{align*}

Again, our goal is to maximize this expression or minimize its opposite:

$$-\log{P(D\vert \theta)} + KL(q_{\lambda}(\theta\vert D)\Vert P(\theta))$$

One step further, we obtain:

\begin{align*} &= -\log{P(D\vert \theta)} + q_{\lambda}(\theta\vert D)\log{q_{\lambda}(\theta\vert D)} - q_{\lambda}(\theta\vert D)\log{P(\theta)}\\ &= \mathop{\mathbb{E}}_{q_{\lambda}(\theta\vert D)}[-\log{P(D\vert \theta)} +\log{q_{\lambda}(\theta\vert D)} - \log{P(\theta)}]\\ &= \mathop{\mathbb{E}}_{q_{\lambda}(\theta\vert D)}[-\big(\log{P(D, \theta)} -\log{q_{\lambda}(\theta\vert D)}\big)]\\ &= -\mathop{\mathbb{E}}_{q_{\lambda}(\theta\vert D)}[\log{P(D, \theta)}] + \mathop{\mathbb{E}}_{q_{\lambda}(\theta\vert D)}[\log{q_{\lambda}(\theta\vert D)}]\\ \end{align*}

In machine learning parlance: "minimize the negative log joint probability of our data and parameter $$\theta$$ — a MAP estimate — plus the entropy of our variational approximation." As a higher entropy is desirable — an approximation which distributes its mass in a conservative fashion — this minimization is a balancing act between the two terms.

For a more in-depth discussion of both entropy and KL-divergence please see Minimizing the Negative Log-Likelihood, in English.

# Posterior predictive distribution

With our estimate for $$\theta$$ as a full distribution, we can now make a new prediction as a full distribution as well.

\begin{align*} P(y\vert x, D) &= \int P(y\vert x, D, \theta)P(\theta\vert x, D)d\theta\\ &= \int P(y\vert x, \theta)P(\theta\vert D)d\theta\\ \end{align*}
• The right term under the integral is the posterior distribution of our parameter $$\theta$$ given the "training" data, $$P(\theta\vert D)$$. Since it does not depend on a new input $$x$$ we have removed $$x$$.
• The left term under the integral is our likelihood function: given an $$x$$ and a $$\theta$$, it produces a $$y$$. While this function does depend on $$\theta$$ — whose values are pulled from our posterior $$P(\theta\vert D)$$ — it does not depend on $$D$$ itself. As such, we have removed $$D$$.

Integrating over $$\theta$$ yields a distribution over $$y$$: we've now captured not just the uncertainty in inference, but also the corresponding uncertain in our predictions.

# What do these distributions actually do for me?

Said differently, "why is it important to quantify uncertainty?"

I think we, as humans, are exceptionally qualified to answer this question: we need to look no further than ourselves, our choices, our environment.

• The cross-walk says "go." Do I:
• Close my eyes, lie down for a 15-second nap in the middle of the road, then walk backwards the rest of the way?
• Quickly look both ways then walk leisurely across the road, keeping an eye out for cyclists at the same time.
• A company emails to say "we'd like to discuss the possibility of a full-time role." Do I:
• Respond saying "Great! Let's chat further" while continuing to speak with other companies.
• Respond saying "Great! Let's chat further" and promptly sever all contact with other companies.
• An extremely reliable lifelong friend calls to say they've found me a beautiful studio in Manhattan for $600/month, and would need to confirm in the next 24 hours if I'd like to take it. Do I: • Take it. • Call three friends to ask if they think that this makes sense. • An extremely sketchy real estate broker calls to say they've found me a beautiful studio in Manhattan for$600/month, and would need to confirm in the next 24 hours if I'd like to take it. Do I:
• Take it.
• Call three friends to ask if they think that this makes sense.

The notion is the same in probabilistic modeling. Furthermore, we often build models with "not big data," and therefore have a substantially non-zero amount of uncertainty in our parameter estimates and subsequent predictions.

Finally, with distributional estimates in hand, we can begin to make more robust, measured and logical decisions. We can do this because, while point estimates give a quick summary of the dynamics of our system, distributions tell the full, thorough story: where the peaks are, their width and height, their distance from one another, etc. For an excellent exploration of what we can do with posterior distributions, check out Rasmus Bååth's Probable Points and Credible Intervals, Part 2: Decision Theory.

Many thanks for reading, and to our pool ring Bayes'.