minor changes

lectures/102-Varada-lectures/class_demos/demo_04-kNNs-SVMs.ipynb

@@ -2916,6 +2916,14 @@
     "for idx in test_idx: \n",
     "    show_nearest_neighbors(idx, nn, Z_valid, X_valid, y_valid)"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "9961525a-1dfe-4f55-a212-18d7bf4460bb",
+   "metadata": {},
+   "source": [
+    "We can use $k$-NNs for more than just classifying images; we can also find the most similar examples in a dataset. You can see how this would be useful in recommendation systems. For instance, if a user is purchasing an item, you can find similar items in the dataset and recommend them to the user! "
+   ]
   }
  ],
  "metadata": {