File size: 18,872 Bytes
513edc0
 
 
 
 
580fc95
4dc9c64
513edc0
 
 
 
 
 
 
4dc9c64
 
 
 
513edc0
 
 
 
 
 
 
 
 
 
 
4dc9c64
513edc0
 
4dc9c64
513edc0
 
 
 
 
4dc9c64
513edc0
4dc9c64
 
 
 
 
513edc0
 
4dc9c64
 
 
 
 
 
 
513edc0
4dc9c64
513edc0
 
 
 
 
4dc9c64
513edc0
4dc9c64
 
 
 
 
 
 
 
 
 
513edc0
 
4dc9c64
 
513edc0
 
4dc9c64
 
 
513edc0
4dc9c64
513edc0
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
4dc9c64
513edc0
 
 
4dc9c64
513edc0
 
4dc9c64
513edc0
 
 
 
 
4dc9c64
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
513edc0
 
4dc9c64
513edc0
 
 
4dc9c64
513edc0
4dc9c64
513edc0
4dc9c64
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
513edc0
 
 
 
 
 
 
 
 
 
 
 
4dc9c64
 
513edc0
 
 
 
 
 
 
 
 
 
 
4dc9c64
 
 
 
 
 
513edc0
 
4dc9c64
 
 
 
 
 
 
 
513edc0
 
4dc9c64
 
 
 
513edc0
 
4dc9c64
 
 
 
 
 
 
 
 
 
 
 
513edc0
4dc9c64
 
 
513edc0
4dc9c64
 
513edc0
 
 
4dc9c64
513edc0
 
 
 
 
 
 
 
 
4dc9c64
513edc0
4dc9c64
 
 
513edc0
 
 
 
 
4dc9c64
513edc0
4dc9c64
 
 
513edc0
4dc9c64
 
 
 
 
 
 
 
 
 
 
513edc0
 
4dc9c64
 
 
 
 
 
513edc0
4dc9c64
 
 
513edc0
 
4dc9c64
513edc0
 
 
 
4dc9c64
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
513edc0
4dc9c64
513edc0
 
 
 
4dc9c64
 
 
 
513edc0
 
 
 
 
 
4dc9c64
 
 
 
 
 
 
 
 
 
 
 
 
 
 
513edc0
 
 
 
 
 
 
 
 
 
 
 
 
 
4dc9c64
513edc0
4dc9c64
 
 
 
 
513edc0
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
import gradio as gr
import pandas as pd
import numpy as np
import joblib
from model_wrapper import FraudDetectionModel
from preprocessor import FraudDataPreprocessor
from feature_utils import fill_missing_features
import os

# Initialize the fraud detection model
fraud_model = FraudDetectionModel()

# Load model if files exist
try:
    # First, ensure the FraudDataPreprocessor class is available
    import sys
    sys.modules['__main__'].FraudDataPreprocessor = FraudDataPreprocessor
    
    # Load the specific XGBoost model files from your training
    model_path = "fraud_detection_model_xgboost_20250727_145448.joblib"
    preprocessor_path = "preprocessor_20250727_145448.joblib"
    metadata_path = "model_metadata_20250727_145448.joblib"
    
    if os.path.exists(model_path) and os.path.exists(preprocessor_path):
        if os.path.exists(metadata_path):
            fraud_model.load_model(model_path, preprocessor_path, metadata_path)
        else:
            fraud_model.load_model(model_path, preprocessor_path)
        model_loaded = True
        print(f"βœ… Model loaded successfully!")
    else:
        model_loaded = False
        print("❌ Model files not found. Please upload the following files:")
        print("- fraud_detection_model_xgboost_20250727_145448.joblib")
        print("- preprocessor_20250727_145448.joblib") 
        print("- model_metadata_20250727_145448.joblib")
except Exception as e:
    model_loaded = False
    print(f"❌ Error loading model: {e}")

def predict_single_transaction(
    transaction_id,
    transaction_dt,
    transaction_amt,
    product_cd,
    card1,
    card2,
    card3,
    card4,
    card5,
    card6,
    addr1,
    addr2,
    p_emaildomain
):
    """Predict fraud risk for a single transaction with exact API fields"""
    
    if not model_loaded:
        return "❌ Model not loaded. Please contact administrator.", "", "", ""
    
    try:
        # Prepare transaction data exactly as API expects
        transaction_data = {
            'TransactionID': int(transaction_id) if transaction_id else 123456,
            'TransactionDT': int(transaction_dt) if transaction_dt else 18403200,
            'TransactionAmt': float(transaction_amt),
            'ProductCD': product_cd,
            'card1': int(card1) if card1 else None,
            'card2': float(card2) if card2 else None,
            'card3': float(card3) if card3 else None,
            'card4': card4,
            'card5': float(card5) if card5 else None,
            'card6': card6,
            'addr1': float(addr1) if addr1 else None,
            'addr2': float(addr2) if addr2 else None,
            'P_emaildomain': p_emaildomain,
            'R_emaildomain': p_emaildomain  # Often same as P_emaildomain
        }
        
        # Fill missing features with defaults
        complete_data = fill_missing_features(transaction_data)
        
        # Make prediction
        result = fraud_model.predict_single_transaction(complete_data)
        
        if 'error' in result:
            return f"❌ {result['error']}", "", "", ""
        
        # Format results
        probability = result['fraud_probability']
        risk_level = result['risk_level']
        recommendation = result['recommendation']
        
        # Create risk indicator
        if probability >= 0.8:
            risk_indicator = f"πŸ”΄ HIGH RISK ({probability:.1%})"
        elif probability >= 0.5:
            risk_indicator = f"🟑 MEDIUM RISK ({probability:.1%})"
        elif probability >= 0.2:
            risk_indicator = f"🟠 LOW RISK ({probability:.1%})"
        else:
            risk_indicator = f"🟒 VERY LOW RISK ({probability:.1%})"
        
        return risk_indicator, f"{probability:.4f}", risk_level, recommendation
        
    except Exception as e:
        return f"❌ Error: {str(e)}", "", "", ""

def predict_batch_from_csv(file):
    """Predict fraud risk for multiple transactions from CSV"""
    
    if not model_loaded:
        return "❌ Model not loaded. Please contact administrator.", None
    
    if file is None:
        return "❌ Please upload a CSV file.", None
    
    try:
        # Read CSV file
        df = pd.read_csv(file.name)
        
        # Validate required columns
        required_cols = ['TransactionAmt']
        missing_cols = [col for col in required_cols if col not in df.columns]
        
        if missing_cols:
            return f"❌ Missing required columns: {missing_cols}. Please ensure your CSV has at least 'TransactionAmt' column.", None
        
        # Add default TransactionID if not present
        if 'TransactionID' not in df.columns:
            df['TransactionID'] = range(1, len(df) + 1)
        
        # Process each row and make predictions
        results = []
        
        for idx, row in df.iterrows():
            try:
                # Fill missing features for this row
                transaction_data = row.to_dict()
                complete_data = fill_missing_features(transaction_data)
                
                # Make prediction
                result = fraud_model.predict_single_transaction(complete_data)
                
                if 'error' not in result:
                    # Add results to original row data
                    row_result = row.copy()
                    row_result['fraud_probability'] = result['fraud_probability']
                    row_result['risk_level'] = result['risk_level']
                    row_result['recommendation'] = result['recommendation']
                    row_result['is_suspicious'] = result['is_suspicious']
                else:
                    # Handle prediction error
                    row_result = row.copy()
                    row_result['fraud_probability'] = None
                    row_result['risk_level'] = 'Error'
                    row_result['recommendation'] = result.get('error', 'Prediction failed')
                    row_result['is_suspicious'] = False
                
                results.append(row_result)
                
            except Exception as e:
                # Handle row processing error
                row_result = row.copy()
                row_result['fraud_probability'] = None
                row_result['risk_level'] = 'Error'
                row_result['recommendation'] = f'Processing error: {str(e)}'
                row_result['is_suspicious'] = False
                results.append(row_result)
        
        # Create results DataFrame
        results_df = pd.DataFrame(results)
        
        # Save results
        output_path = "fraud_predictions_batch.csv"
        results_df.to_csv(output_path, index=False)
        
        # Create summary
        valid_predictions = results_df[results_df['fraud_probability'].notna()]
        total_transactions = len(results_df)
        valid_count = len(valid_predictions)
        
        if valid_count > 0:
            high_risk = len(valid_predictions[valid_predictions['fraud_probability'] >= 0.8])
            medium_risk = len(valid_predictions[(valid_predictions['fraud_probability'] >= 0.5) & (valid_predictions['fraud_probability'] < 0.8)])
            low_risk = len(valid_predictions[(valid_predictions['fraud_probability'] >= 0.2) & (valid_predictions['fraud_probability'] < 0.5)])
            very_low_risk = len(valid_predictions[valid_predictions['fraud_probability'] < 0.2])
            
            summary = f"""
πŸ“Š **Batch Prediction Summary**

Total Transactions: {total_transactions}
Successfully Processed: {valid_count}
Errors: {total_transactions - valid_count}

**Risk Distribution:**
πŸ”΄ High Risk: {high_risk} ({high_risk/valid_count:.1%})
🟑 Medium Risk: {medium_risk} ({medium_risk/valid_count:.1%})
🟠 Low Risk: {low_risk} ({low_risk/valid_count:.1%})
🟒 Very Low Risk: {very_low_risk} ({very_low_risk/valid_count:.1%})

Results saved to: {output_path}
            """
        else:
            summary = f"""
❌ **Batch Processing Failed**

Total Transactions: {total_transactions}
Successfully Processed: 0
All transactions encountered errors.

Please check your CSV format and try again.
            """
        
        return summary, output_path
        
    except Exception as e:
        return f"❌ Error processing CSV: {str(e)}", None

# Create Gradio interface
with gr.Blocks(title="Fraud Detection System", theme=gr.themes.Soft()) as app:
    
    gr.Markdown("""
    # πŸ”’ Credit Card Fraud Detection System
    
    This system uses **XGBoost machine learning** to assess the risk of credit card transactions being fraudulent.
    Enter transaction details for single prediction or upload CSV for batch processing.
    
    **Risk Levels:**
    - πŸ”΄ High Risk (β‰₯80%): Block transaction immediately
    - 🟑 Medium Risk (50-79%): Manual review required  
    - 🟠 Low Risk (20-49%): Monitor transaction
    - 🟒 Very Low Risk (<20%): Process normally
    """)
    
    with gr.Tabs():
        
        # Single Transaction Tab
        with gr.TabItem("πŸ” Single Transaction"):
            gr.Markdown("""
            ### Single Transaction Fraud Detection
            Enter the transaction details below for instant fraud risk assessment.
            """)
            
            with gr.Row():
                with gr.Column():
                    gr.Markdown("### πŸ“ Transaction Information")
                    transaction_id = gr.Number(label="Transaction ID", value=123456, precision=0)
                    transaction_dt = gr.Number(label="Transaction DateTime (seconds)", value=18403200, precision=0)
                    transaction_amt = gr.Number(label="Transaction Amount ($)", value=150.00)
                    product_cd = gr.Dropdown(
                        choices=["W", "C", "S", "R", "H"],
                        label="Product Code",
                        value="W"
                    )
                    
                    gr.Markdown("### πŸ’³ Card Information")
                    card1 = gr.Number(label="Card 1", value=4532015112830366, precision=0)
                    card2 = gr.Number(label="Card 2", value=404.0)
                    card3 = gr.Number(label="Card 3", value=150.0)
                
                with gr.Column():
                    gr.Markdown("### πŸ’³ Card Details")
                    card4 = gr.Dropdown(
                        choices=["visa", "mastercard", "american express", "discover"],
                        label="Card Type",
                        value="visa"
                    )
                    card5 = gr.Number(label="Card 5", value=142.0)
                    card6 = gr.Dropdown(
                        choices=["credit", "debit"],
                        label="Card Category",
                        value="credit"
                    )
                    
                    gr.Markdown("### πŸ“ Address Information") 
                    addr1 = gr.Number(label="Address 1", value=315.0)
                    addr2 = gr.Number(label="Address 2", value=87.0)
                    
                    gr.Markdown("### πŸ“§ Email Information")
                    p_emaildomain = gr.Textbox(label="Email Domain", value="gmail.com")
            
            predict_btn = gr.Button("πŸ” Analyze Transaction", variant="primary", size="lg")
            
            gr.Markdown("### πŸ“Š Prediction Results")
            with gr.Row():
                risk_output = gr.Textbox(label="Risk Assessment", lines=1)
                probability_output = gr.Textbox(label="Fraud Probability", lines=1)
                
            with gr.Row():
                risk_level_output = gr.Textbox(label="Risk Level", lines=1)
                recommendation_output = gr.Textbox(label="Recommendation", lines=2)
            
            predict_btn.click(
                predict_single_transaction,
                inputs=[
                    transaction_id, transaction_dt, transaction_amt, product_cd,
                    card1, card2, card3, card4, card5, card6,
                    addr1, addr2, p_emaildomain
                ],
                outputs=[risk_output, probability_output, risk_level_output, recommendation_output]
            )
        
        # Batch Processing Tab
        with gr.TabItem("πŸ“ Batch Processing"):
            gr.Markdown("""
            ### CSV Batch Processing
            
            Upload a CSV file containing multiple transactions for batch fraud detection.
            
            **Required CSV Columns:**
            - `TransactionAmt` (required)
            - `TransactionID` (optional - will be auto-generated)
            - `TransactionDT`, `ProductCD`, `card1-6`, `addr1-2`, `P_emaildomain` (optional - smart defaults used)
            
            **Example CSV Format:**
            ```
            TransactionID,TransactionDT,TransactionAmt,ProductCD,card1,card2,card3,card4,card5,card6,addr1,addr2,P_emaildomain
            123456,18403200,150.00,W,4532015112830366,404.0,150.0,visa,142.0,credit,315.0,87.0,gmail.com
            123457,18403300,2500.00,C,5555555555554444,555.0,200.0,mastercard,224.0,credit,420.0,95.0,yahoo.com
            ```
            """)
            
            file_upload = gr.File(
                label="Upload CSV File", 
                file_types=[".csv"],
                elem_id="csv-upload"
            )
            batch_btn = gr.Button("πŸ” Process Batch", variant="primary", size="lg")
            
            gr.Markdown("### πŸ“Š Batch Results")
            batch_output = gr.Textbox(label="Processing Summary", lines=12)
            download_file = gr.File(label="Download Results CSV")
            
            batch_btn.click(
                predict_batch_from_csv,
                inputs=[file_upload],
                outputs=[batch_output, download_file]
            )
        
        # Sample Data Tab
        with gr.TabItem("πŸ“‹ Sample Data"):
            gr.Markdown("""
            ### Sample Transaction Data
            
            Use these examples to test the system or as a template for your CSV files.
            """)
            
            gr.Markdown("""
            #### Example 1: Low Risk Transaction
            ```json
            {
                "TransactionID": 123456,
                "TransactionDT": 18403200,
                "TransactionAmt": 150.00,
                "ProductCD": "W",
                "card1": 4532015112830366,
                "card2": 404.0,
                "card3": 150.0,
                "card4": "visa",
                "card5": 142.0,
                "card6": "credit",
                "addr1": 315.0,
                "addr2": 87.0,
                "P_emaildomain": "gmail.com"
            }
            ```
            
            #### Example 2: Higher Risk Transaction
            ```json
            {
                "TransactionID": 123457,
                "TransactionDT": 18403300,
                "TransactionAmt": 2500.00,
                "ProductCD": "C",
                "card1": 5555555555554444,
                "card2": 555.0,
                "card3": 200.0,
                "card4": "mastercard",
                "card5": 224.0,
                "card6": "credit",
                "addr1": 420.0,
                "addr2": 95.0,
                "P_emaildomain": "yahoo.com"
            }
            ```
            
            #### CSV Sample File
            You can copy this into a CSV file for batch testing:
            ```
            TransactionID,TransactionDT,TransactionAmt,ProductCD,card1,card2,card3,card4,card5,card6,addr1,addr2,P_emaildomain
            123456,18403200,150.00,W,4532015112830366,404.0,150.0,visa,142.0,credit,315.0,87.0,gmail.com
            123457,18403300,2500.00,C,5555555555554444,555.0,200.0,mastercard,224.0,credit,420.0,95.0,yahoo.com
            123458,18403400,75.50,W,4111111111111111,300.0,75.0,visa,100.0,debit,200.0,50.0,hotmail.com
            ```
            """)
        
        # Model Info Tab
        with gr.TabItem("ℹ️ Model Information"):
            if model_loaded and fraud_model.metadata:
                model_info = fraud_model.get_model_info()
                gr.Markdown(f"""
                ### Model Status
                **Status:** βœ… {model_info.get('model_name', 'XGBoost')} Model Loaded  
                **AUC Score:** {model_info.get('auc_score', 'N/A')}  
                **Training Date:** {model_info.get('training_timestamp', 'N/A')}  
                **Features:** {model_info.get('feature_count', 'N/A')}  
                
                ### About This Model
                This fraud detection system uses an **XGBoost classifier** trained on a comprehensive dataset 
                of credit card transactions. The model achieved high performance with advanced feature engineering
                and ensemble learning techniques.
                
                ### API Compatible Interface
                This interface matches the exact field structure expected by the fraud detection API:
                
                **Single Prediction Endpoint:** `/v1/predict`  
                **Batch Prediction Endpoint:** `/v1/predict/batch`
                
                ### Supported Fields
                - **TransactionID**: Unique transaction identifier
                - **TransactionDT**: Transaction datetime (seconds)
                - **TransactionAmt**: Transaction amount in USD
                - **ProductCD**: Product code (W, C, S, R, H)
                - **card1-6**: Card-related features
                - **addr1-2**: Address information
                - **P_emaildomain**: Primary email domain
                
                ### Model Performance
                - **Algorithm**: XGBoost (Extreme Gradient Boosting)
                - **AUC Score**: {model_info.get('auc_score', 'N/A')}
                - **Features Used**: {model_info.get('feature_count', 'N/A')} engineered features
                - **Training Method**: Cross-validation with stratified sampling
                - **Speed**: Real-time predictions (<100ms)
                """)
            else:
                gr.Markdown(f"""
                ### Model Status
                **Status:** {'βœ… Basic Model Loaded' if model_loaded else '❌ Not Loaded'}
                
                ### About This Model
                This fraud detection system uses advanced machine learning algorithms to assess transaction risk.
                The model processes transactions with the same field structure as the API endpoints.
                
                ### Features
                - Single transaction analysis
                - Batch CSV processing
                - Real-time risk assessment
                - API-compatible field structure
                """)

# Launch the app
if __name__ == "__main__":
    app.launch()