Update app.py

app.py

@@ -14,7 +14,10 @@ model_card = f"""
 ## Description
 
 The **DecisionTreeClassifier** employs a pruning technique that can be configured using the cost complexity parameter, commonly referred to as **ccp_alpha**.
-By increasing the value of **ccp_alpha**, a greater number of nodes can be pruned. This demo demonstrates the impact of **ccp_alpha** on tree regularization
+By increasing the value of **ccp_alpha**, a greater number of nodes can be pruned.
+In this demo, a DecisionTreeClassifier will be trained on the Breast Cancer dataset. Then, the effect of **ccp_alpha** in many terms of the tree-based model like the impurity of leaves, depth, number of nodes, and accuracy on train and test data are shown in many figures.
+Based on this information, the best number of **ccp_alpha** is chosen. This demo also shows the results of the best **ccp_alpha** with accuracy on train and test datasets.
+You can play around with different ``test size`` and  ``random state``
 
 ## Dataset
 
@@ -54,7 +57,6 @@ def get_ccp(test_size, random_state):
     ax2.set_title("Number of nodes vs alpha")
 
     fig3, ax3 = plt.subplots()
-
     ax3.plot(ccp_alphas, depth, marker="o", drawstyle="steps-post")
     ax3.set_xlabel("alpha")
     ax3.set_ylabel("depth of tree")