Update src/streamlit_app.py

src/streamlit_app.py

@@ -35,21 +35,20 @@ else:
 # ================= TITLE =================
 st.title("💳 Credit Card Fraud Detection")
 st.markdown("Real-time fraud detection using Machine Learning.")
+
 st.markdown("""
 ### 🧠 How the model works
-
 The model evaluates each transaction using three types of signals:
 
 - ⏱ **Time** – seconds since the first recorded transaction  
 - 💰 **Amount** – transaction value  
 - 🔐 **V1–V28** – anonymized PCA components capturing hidden behavioral patterns  
 
-These components do not represent directly interpretable features.  
-They are mathematical transformations of the original transaction data used to protect sensitive financial information while preserving the patterns needed for fraud detection.
+These components are mathematical transformations of the original data and are not directly interpretable.
 """)
 
 tab1, tab2 = st.tabs(["🔍 Prediction", "📊 Model Insights"])
-st.info("V1–V28 are PCA-transformed features. They are not directly interpretable but are highly predictive for fraud detection.")
+
 # ================= TAB 1 – PREDICTION =================
 with tab1:
 
@@ -69,15 +68,14 @@ with tab1:
         with col2:
             amount = st.number_input("Amount", value=0.0)
 
-        pca_values = []
-        
-st.info("""
+        st.info("""
 V1–V28 are PCA-transformed features.
-
 The original transaction variables are anonymized for privacy.
 These components capture hidden transaction patterns used by the model to detect fraud.
 """)
-        
+
+        pca_values = []
+
         with st.expander("PCA Features (Anonymized Transaction Patterns)"):
             for i in range(28):
                 val = st.number_input(f"V{i+1}", value=0.0)
@@ -114,20 +112,16 @@ with tab2:
         st.info("Test data not found. Upload X_test.npy and y_test.npy to enable insights.")
 
     else:
-
         st.metric("AUC", round(auc, 3))
 
-        # ROC Curve
         fig, ax = plt.subplots()
         RocCurveDisplay.from_predictions(y_test, y_prob, ax=ax)
         st.pyplot(fig)
 
-        # Confusion Matrix
         fig2, ax2 = plt.subplots()
         ConfusionMatrixDisplay.from_predictions(y_test, y_pred, ax=ax2)
         st.pyplot(fig2)
 
-    # Feature importance (tree-based models only)
     if hasattr(model, "feature_importances_"):
 
         st.subheader("Feature Importance")