Update src/streamlit_app.py

src/streamlit_app.py

@@ -13,10 +13,9 @@ st.set_page_config(page_title="FinTech Fraud Guard", layout="wide")
 st.markdown("""
     <style>
     .main { background-color: #0e1117; }
-    .stButton>button { width: 100%; border-radius: 20px; background: linear-gradient(45deg, #ff4b4b, #ff8f8f); color: white; border: none; font-weight: bold; }
-    .fraud-card { padding: 25px; border-radius: 15px; background-color: #1e2130; border-left: 8px solid #ff4b4b; margin-bottom: 20px; box-shadow: 0 4px 15px rgba(255, 75, 75, 0.2); }
-    .legit-card { padding: 25px; border-radius: 15px; background-color: #1e2130; border-left: 8px solid #00ffcc; margin-bottom: 20px; box-shadow: 0 4px 15px rgba(0, 255, 204, 0.1); }
-    .stTextInput>div>div>input, .stNumberInput>div>div>input { background-color: #262730; color: white; border-radius: 10px; }
+    .stButton>button { width: 100%; border-radius: 20px; background: linear-gradient(45deg, #ff4b4b, #ff8f8f); color: white; border: none; }
+    .fraud-card { padding: 20px; border-radius: 15px; background-color: #1e2130; border-left: 5px solid #ff4b4b; margin-bottom: 20px; }
+    .legit-card { padding: 20px; border-radius: 15px; background-color: #1e2130; border-left: 5px solid #00ffcc; margin-bottom: 20px; }
     </style>
     """, unsafe_allow_html=True)
 
@@ -103,100 +102,100 @@ class MoEFraudModel(nn.Module):
         h_final = (weights[:, 0:1]*h1 + weights[:, 1:2]*h2 + weights[:, 2:3]*h3 + weights[:, 3:4]*h4)
         return self.classifier(h_final)
 
+def haversine(lat1, lon1, lat2, lon2):
+    r = 6371
+    phi1, phi2 = np.radians(lat1), np.radians(lat2)
+    dphi = np.radians(lat2 - lat1)
+    dlambda = np.radians(lon2 - lon1)
+    a = np.sin(dphi/2)**2 + np.cos(phi1)*np.cos(phi2)*np.sin(dlambda/2)**2
+    return 2 * r * np.arcsin(np.sqrt(a))
+
 REPO_ID = "rocky250/FinTech"
 
 @st.cache_resource
 def load_assets():
-    weights_path = hf_hub_download(repo_id=REPO_ID, filename="proposed_moe_model.pth")
-    config_path = hf_hub_download(repo_id=REPO_ID, filename="config.json")
-    encoders_path = hf_hub_download(repo_id=REPO_ID, filename="label_encoders.joblib")
-    scaler_path = hf_hub_download(repo_id=REPO_ID, filename="scaler.joblib")
-    with open(config_path, 'r') as f:
+    w_p = hf_hub_download(repo_id=REPO_ID, filename="proposed_moe_model.pth")
+    c_p = hf_hub_download(repo_id=REPO_ID, filename="config.json")
+    e_p = hf_hub_download(repo_id=REPO_ID, filename="label_encoders.joblib")
+    s_p = hf_hub_download(repo_id=REPO_ID, filename="scaler.joblib")
+    with open(c_p, 'r') as f:
         config = json.load(f)
     model = MoEFraudModel(config['cat_dims'], config['num_cols_map'], config['embed_dim'])
-    model.load_state_dict(torch.load(weights_path, map_location=torch.device('cpu')))
+    model.load_state_dict(torch.load(w_p, map_location=torch.device('cpu')))
     model.eval()
-    return model, joblib.load(encoders_path), joblib.load(scaler_path), config
+    return model, joblib.load(e_p), joblib.load(s_p), config
 
 model, encoders, scaler, config = load_assets()
 
-st.title("FinTech Mixture-of-Experts Fraud Guard")
-st.markdown("---")
+st.title("FinTech Fraud Analysis")
+st.markdown("### MoE Transaction Verifier")
 
-with st.form("main_form"):
-    c1, c2 = st.columns(2)
-    
+with st.form("transaction_form"):
+    c1, c2, c3 = st.columns(3)
     with c1:
-        st.subheader("Transaction Details")
-        trans_dt = st.text_input("Transaction Date & Time", value="21-06-2020 12:14")
-        amt = st.number_input("Transaction Amount ($)", value=2.86, format="%.2f")
-        cc_num = st.text_input("Credit Card Number", "2.29116E+15")
-        merchant = st.text_input("Merchant Name", value="fraud_Kirlin and Sons")
-        category = st.text_input("Category", value="personal_care")
-        
-    with c2:
-        st.subheader("Cardholder Info")
-        col_a, col_b = st.columns(2)
-        first_name = col_a.text_input("First Name", value="Jeff")
-        last_name = col_b.text_input("Last Name", value="Elliott")
+        st.subheader("Personal & Card")
+        first = st.text_input("First Name", "Jeff")
+        last = st.text_input("Last Name", "Elliott")
         gender = st.selectbox("Gender", ["M", "F"])
-        street = st.text_input("Street Address", value="351 Darlene Green")
-
-    submit = st.form_submit_button("RUN FRAUD ANALYSIS")
+        dob = st.text_input("Date of Birth (DD-MM-YYYY)", "19-03-1968")
+        cc_num = st.text_input("CC Number", "2.29E+15")
+        job = st.text_input("Job", "Mechanical engineer")
+    with c2:
+        st.subheader("Transaction Details")
+        trans_dt = st.text_input("Transaction Time (DD-MM-YYYY HH:MM)", "21-06-2020 12:14")
+        merchant = st.text_input("Merchant", "fraud_Kirlin and Sons")
+        category = st.text_input("Category", "personal_care")
+        amt = st.number_input("Amount ($)", value=2.86)
+        u_time = st.number_input("Unix Time", value=1371816865)
+    with c3:
+        st.subheader("Geography")
+        street = st.text_input("Street", "351 Darlene Green")
+        city = st.text_input("City", "Columbia")
+        state = st.text_input("State", "SC")
+        zip_v = st.text_input("Zip Code", "29209")
+        city_pop = st.number_input("City Population", value=333497)
+        lat = st.number_input("Lat", value=33.9659)
+        lon = st.number_input("Long", value=-80.9355)
+        m_lat = st.number_input("Merch Lat", value=33.986391)
+        m_lon = st.number_input("Merch Long", value=-81.200714)
+
+    submit = st.form_submit_button("ANALYZE TRANSACTION")
 
 if submit:
     try:
         dt_obj = datetime.strptime(trans_dt, "%d-%m-%Y %H:%M")
+        dob_obj = datetime.strptime(dob, "%d-%m-%Y")
+        age = (dt_obj - dob_obj).days // 365
+        dist = haversine(lat, lon, m_lat, m_lon)
         
         num_feats = [
-            amt, dt_obj.hour, dt_obj.weekday(), 
-            1 if dt_obj.weekday() >= 5 else 0, dt_obj.timestamp(),
-            50000, 35, 3600, 50.0, 10.0, 100.0, 
-            0.02, 1000, 33.5, -86.8, 33.6, -86.9, 15.5, 0
+            amt, dt_obj.hour, dt_obj.weekday(), 1 if dt_obj.weekday() >= 5 else 0, u_time,
+            city_pop, age, 3600, 50.0, 15.0, 150.0, 0.01, 500, lat, lon, m_lat, m_lon, dist, 0
         ]
         
-        cat_order = ['category', 'cc_num', 'gender', 'job', 'city', 'state', 'zip', 'merchant']
-        cat_vals = [category, cc_num, gender, job, city, state, zip_code, merchant]
+        cat_feats = [category, cc_num, gender, job, city, state, zip_v, merchant]
         
         num_input = scaler.transform([num_feats])
-        
         cat_encoded = []
-        for i, col_name in enumerate(cat_order):
-            val = str(cat_vals[i])
-            if val in encoders[col_name].classes_:
-                cat_encoded.append(encoders[col_name].transform([val])[0])
+        for i, col in enumerate(config['cat_dims'].keys()):
+            val = str(cat_feats[i])
+            if val in encoders[col].classes_:
+                cat_encoded.append(encoders[col].transform([val])[0])
             else:
                 cat_encoded.append(0)
         
         with torch.no_grad():
             cat_t = torch.tensor([cat_encoded], dtype=torch.long)
             num_t = torch.tensor(num_input, dtype=torch.float32)
-            logits = model(cat_t, num_t)
-            prob = torch.sigmoid(logits).item()
+            prob = torch.sigmoid(model(cat_t, num_t)).item()
 
-        st.markdown("### Analysis Result")
+        st.divider()
         if prob > 0.5:
-            st.markdown(f"""
-                <div class="fraud-card">
-                    <h2 style="color:#ff4b4b; margin-top:0;">High Risk Detected</h2>
-                    <p style="font-size:1.2rem;">Fraud Probability: <strong>{prob*100:.2f}%</strong></p>
-                    <hr style="border:0.5px solid #ff4b4b33;">
-                    <p>The MoE Gating Network has flagged this transaction as highly suspicious based on merchant patterns and transaction timing.</p>
-                </div>
-                """, unsafe_allow_html=True)
+            st.markdown(f'<div class="fraud-card"><h2>High Risk Detected!</h2><p>Confidence: {prob*100:.2f}%</p></div>', unsafe_allow_html=True)
         else:
-            st.markdown(f"""
-                <div class="legit-card">
-                    <h2 style="color:#00ffcc; margin-top:0;">Transaction Verified</h2>
-                    <p style="font-size:1.2rem;">Fraud Probability: <strong>{prob*100:.2f}%</strong></p>
-                    <hr style="border:0.5px solid #00ffcc33;">
-                    <p>Our experts have analyzed the spatial and behavioral features and determined this transaction is likely legitimate.</p>
-                </div>
-                """, unsafe_allow_html=True)
+            st.markdown(f'<div class="legit-card"><h2>Transaction Safe</h2><p>Confidence: {(1-prob)*100:.2f}%</p></div>', unsafe_allow_html=True)
                 
     except Exception as e:
-        st.error(f"Validation Error: Ensure Date Format is DD-MM-YYYY HH:MM. Raw Error: {e}")
+        st.error(f"Error: {e}")
 
-st.sidebar.title("System Info")
-st.sidebar.info("Model: MoE Neural Network")
-st.sidebar.write("Architecture: 4 Experts (Temporal, Behavioral, Frequency, Spatial)")
+st.sidebar.info("MoE model utilizing 4 expert networks for fraud detection.")