Soften "estimation is optimization" to "often" in deck and ch1

Author: bnaras

01-intro.qmd

@@ -18,7 +18,7 @@ library(ggplot2)
 
 ## Optimization is everywhere in statistics
 
-Estimation is optimization. We **maximize** a likelihood, **minimize** a loss, or
+Estimation is often optimization. We **maximize** a likelihood, **minimize** a loss, or
 **penalize** a fit to trade bias for variance. Ordinary least squares, ridge and
 lasso, logistic regression, support vector machines, maximum-entropy models,
 portfolio selection --- each is "find the parameter that makes some criterion as

slides/intro.qmd

@@ -25,7 +25,7 @@ This is the **opening** to a hands-on session. The real work lives in the
 stage, then we drop into code.
 
 ::: incremental
-- **Estimation is optimization.** You already do this every day.
+- **Estimation is often optimization.** You already do this every day.
 - A huge share of statistics and ML is, underneath, **convex optimization**.
 - **CVXR** lets you write these problems in *natural mathematical notation* and
   have them solved to **global** optimality --- no custom solver, no matching to a