machinelearning

commit 8c10e4a12b13cd54a8af0a73edad20d8b8ffbf5a
parent e23879d5f5e27fbec6e1eec7a78e2a5072ccce72
Author: Andrew <andrewlaack1@gmail.com>
Date:   Wed, 17 Jul 2024 18:03:16 -0500

added polynomial feature logistic regression.

Diffstat:
M
diabetes/DiabetesPrediction.ipynb
 | 
47
+++++++++++++++++++++++++++++++++++++++++++++++

1 file changed, 47 insertions(+), 0 deletions(-)
diff --git a/diabetes/DiabetesPrediction.ipynb b/diabetes/DiabetesPrediction.ipynb
@@ -8,6 +8,13 @@
    ]
   },
   {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "Logistic regression was almost 80% accurate followed closely at 77% by a neural network. Adaboost, Random Forest, and Polynomial feature logistic regression were in the low 70s."
+   ]
+  },
+  {
    "cell_type": "code",
    "execution_count": 374,
    "metadata": {},
@@ -4714,6 +4721,46 @@
     "accuracies.append(\"Neural Network: \" + str(accuracy_score(y_true=y_test,y_pred=y_pred)))\n",
     "accuracies"
    ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 414,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.preprocessing import PolynomialFeatures\n",
+    "\n",
+    "poly = PolynomialFeatures(degree=2)\n",
+    "X_train_poly = poly.fit_transform(X_train)\n",
+    "X_val_poly = poly.transform(X_val)\n",
+    "X_test_poly = poly.transform(X_test)\n",
+    "\n",
+    "\n",
+    "log_reg = LogisticRegression()\n",
+    "log_reg = log_reg.fit(X_train_poly,y_train)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 415,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "0.7604166666666666"
+      ]
+     },
+     "execution_count": 415,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "y_pred = log_reg.predict(X_test_poly)\n",
+    "\n",
+    "accuracy_score(y_true=y_test,y_pred=y_pred)"
+   ]
   }
  ],
  "metadata": {