Changes for dsm020-cw2.ipynb: 10 added lines, 6 removed lines.
Original line number
Diff line number
Diff line
@@ -206,7 +206,9 @@
"cell_type": "code",
"execution_count": 5,
"id": "7293fe1c-b150-4107-8e91-8c7aa0009bd4",
"metadata": {},
"metadata": {
"scrolled": true
},
"outputs": [
{
"name": "stdout",
@@ -1091,7 +1093,7 @@
},
{
"cell_type": "code",
"execution_count": 110,
"execution_count": 34,
"id": "383ef232-6bbd-459a-a0dd-2370e934cb6d",
"metadata": {
"scrolled": true
@@ -1295,7 +1297,9 @@
"\n",
"Following data preprocessing, this study addressed the issue of class imbalance using the Synthetic Minority Over-sampling Technique (SMOTE) (Imbalanced-learn, n.d.). The dataset is divided into training and test sets, with SMOTE applied only to the training set to prevent data leakage. This results in two versions of the training data: one with SMOTE and one without.\n",
"\n",
"Logistic Regression is trained on both versions of the training set using the built-in classification algorithm from Scikit-learn (Scikit-learn, n.d.), with the number of epochs set to the default value of 100 and L2 penalty. For evaluation, only out-of-sample (OOS) results are reported."
"Logistic Regression is trained on both versions of the training set using the built-in classification algorithm from Scikit-learn (Scikit-learn, n.d.), with the number of epochs set to the default value of 100 and L2 penalty. For evaluation, only out-of-sample (OOS) results are reported.\n",
"\n",
"The experiment was conducted locally on a Mac M1 (2020)."
