machinelearning

commit d40d44c47be6e86d8bf6b91cca6c4eeaa554495f
parent db12c8d974d73a1f955784efa73c85e59e058fc5
Author: Andrew <andrewlaack1@gmail.com>
Date:   Mon,  3 Jun 2024 22:17:18 -0500

Ending the day

Diffstat:
M
mnist/MNISTClassification.ipynb
 | 
41
+++++++++++++++++++++++++++++++++++++++++

1 file changed, 41 insertions(+), 0 deletions(-)
diff --git a/mnist/MNISTClassification.ipynb b/mnist/MNISTClassification.ipynb
@@ -565,6 +565,47 @@
     "# This shows that pure false is quite good...\n",
     "cross_val_score(dummy_clf, X_train, y_train_5, cv=3, scoring='accuracy')"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 38,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.model_selection import cross_val_predict\n",
+    "\n",
+    "# Returns predictions from each fold instead of percentages (like cross_val_score would)\n",
+    "y_train_pred = cross_val_predict(sgd_clf, X_train, y_train_5, cv=3)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 39,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "array([[53481,  1098],\n",
+       "       [ 1546,  3875]])"
+      ]
+     },
+     "execution_count": 39,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "# Create confusion matrix based on actual and predicted values\n",
+    "\n",
+    "# As we can see, in retrospect, there are 1098 images that \n",
+    "# were thought to be 5 that are not and 1546 5's that were \n",
+    "# thought to be not 5.\n",
+    "\n",
+    "from sklearn.metrics import confusion_matrix\n",
+    "cm = confusion_matrix(y_train_5, y_train_pred)\n",
+    "cm"
+   ]
   }
  ],
  "metadata": {