File size: 22,522 Bytes
e80d1c2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
<!DOCTYPE html>
<html lang="en" data-theme="dark">
<head>
    <meta charset="UTF-8">
    <meta name="viewport" content="width=device-width, initial-scale=1.0">
    <title>Customer Churn Analyzer | Documentation</title>
    
    <!-- Google Font -->
    <link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;500;600;700;800&display=swap" rel="stylesheet">
    <!-- Font Awesome Icons -->
    <link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.4.0/css/all.min.css">

    <style>
        /* CSS Variable Theme mirroring predict.html */
        :root {
            --bg-color: #f8fafc;         /* Slate 50 - Light, clean background */
            --card-bg: #ffffff;          /* Pure white for form containers */
            --text-main: #0f172a;        /* Slate 900 - Deep, readable text */
            --text-muted: #64748b;       /* Slate 500 - For labels and placeholders */
            --border-color: #cbd5e1;     /* Slate 300 - Subtle input borders */
            --primary-accent: #2563eb;   /* Blue 600 - Focus state & primary buttons */
            --primary-hover: #1d4ed8;    /* Blue 700 */
            --font-family: 'Inter', sans-serif;
            --card-shadow: 0 15px 40px rgba(15, 23, 42, 0.06);
            --card-border-radius: 28px;
            --inner-border-radius: 18px;
            --gradient-start: #1e3a8a;
            --gradient-end: #3b82f6;
            
            --success: #10b981;
            --warning: #f59e0b;
            --danger: #ef4444;
        }

        /* Dark Mode variables */
        [data-theme="dark"] {
            --bg-color: #0f172a;
            --card-bg: #1e293b;
            --text-main: #f8fafc;
            --text-muted: #94a3b8;
            --border-color: #334155;
            --primary-accent: #3b82f6;
            --primary-hover: #60a5fa;
            --card-shadow: 0 15px 40px rgba(0, 0, 0, 0.3);
            --gradient-start: #0f172a;
            --gradient-end: #1e3a8a;
        }

        * {
            margin: 0;
            padding: 0;
            box-sizing: border-box;
        }

        body {
            font-family: var(--font-family);
            background: linear-gradient(to bottom right, var(--bg-color), var(--bg-color));
            min-height: 100vh;
            padding: 40px 20px;
            color: var(--text-main);
            transition: background 0.3s, color 0.3s;
        }

        .container {
            max-width: 1250px;
            margin: auto;
            position: relative;
        }

        /* Dark Mode Switch Button */
        .theme-toggle-btn {
            position: absolute;
            top: 20px;
            right: 20px;
            background: rgba(37, 99, 235, 0.08);
            border: 1px solid var(--border-color);
            color: var(--primary-accent);
            padding: 8px 12px;
            border-radius: 20px;
            cursor: pointer;
            font-size: 0.85rem;
            font-weight: 600;
            display: flex;
            align-items: center;
            gap: 6px;
            transition: all 0.2s;
            z-index: 100;
        }

        .theme-toggle-btn:hover {
            background: var(--primary-accent);
            color: #ffffff;
        }

        /* HERO SECTION */
        .hero-section {
            text-align: center;
            margin-bottom: 50px;
        }

        .hero-badge {
            display: inline-block;
            padding: 10px 18px;
            border-radius: 999px;
            background: rgba(37, 99, 235, 0.08);
            color: var(--primary-accent);
            font-size: 0.85rem;
            font-weight: 700;
            letter-spacing: 0.4px;
            margin-bottom: 22px;
        }

        .hero-section h1 {
            font-size: 3rem;
            font-weight: 800;
            color: var(--text-main);
            margin-bottom: 18px;
            letter-spacing: -0.02em;
        }

        .hero-section p {
            max-width: 850px;
            margin: auto;
            color: var(--text-muted);
            line-height: 1.9;
            font-size: 1.05rem;
        }

        /* MAIN GRID */
        .documentation-grid {
            display: grid;
            grid-template-columns: 1.2fr 0.8fr;
            gap: 28px;
            align-items: start;
        }

        /* GENERAL CARD */
        .card {
            background: var(--card-bg);
            backdrop-filter: blur(14px);
            border-radius: var(--card-border-radius);
            padding: 34px;
            border: 1px solid var(--border-color);
            box-shadow: var(--card-shadow);
            transition: background-color 0.3s, border-color 0.3s;
        }

        /* LARGE CARD */
        .large-card {
            min-height: 100%;
        }

        /* STICKY COLUMN */
        .sticky-column {
            position: sticky;
            top: 30px;
            display: flex;
            flex-direction: column;
            gap: 24px;
        }

        /* SMALL CARD */
        .small-card {
            padding: 28px;
            transition: transform 0.25s ease, background-color 0.3s, border-color 0.3s;
        }

        .small-card:hover {
            transform: translateY(-2px);
        }

        /* CARD TITLE */
        .card-title {
            font-size: 1.45rem;
            font-weight: 700;
            color: var(--text-main);
            margin-bottom: 22px;
            display: flex;
            align-items: center;
            gap: 8px;
        }

        /* PARAGRAPH */
        .card p {
            color: var(--text-muted);
            line-height: 1.9;
            font-size: 0.97rem;
            margin-bottom: 18px;
        }

        /* FEATURE LIST */
        .feature-list {
            display: flex;
            flex-direction: column;
            gap: 18px;
            margin-top: 22px;
        }

        .feature-item {
            background: var(--bg-color);
            border: 1px solid var(--border-color);
            border-radius: var(--inner-border-radius);
            padding: 18px 20px;
            transition: background-color 0.3s, border-color 0.3s;
        }

        .feature-item strong {
            display: block;
            color: var(--text-main);
            margin-bottom: 7px;
            font-size: 0.96rem;
        }

        .feature-item span {
            color: var(--text-muted);
            line-height: 1.7;
            font-size: 0.92rem;
        }

        /* HIGHLIGHT BOX */
        .highlight-box {
            background: linear-gradient(135deg, rgba(37, 99, 235, 0.05), var(--bg-color));
            border: 1px solid var(--border-color);
            border-radius: 22px;
            padding: 24px;
            margin-top: 28px;
        }

        .highlight-title {
            font-size: 1rem;
            font-weight: 700;
            color: var(--primary-accent);
            margin-bottom: 10px;
        }

        .highlight-box p {
            margin-bottom: 0;
            color: var(--text-main);
            opacity: 0.95;
        }

        /* PIPELINE SECTION */
        .pipeline-card {
            margin-top: 32px;
        }

        .pipeline {
            margin-top: 24px;
            display: flex;
            flex-direction: column;
            gap: 18px;
        }

        .pipeline-step {
            display: flex;
            align-items: flex-start;
            gap: 16px;
            background: var(--bg-color);
            border: 1px solid var(--border-color);
            border-radius: var(--inner-border-radius);
            padding: 18px;
            transition: background-color 0.3s, border-color 0.3s;
        }

        .step-number {
            min-width: 38px;
            height: 38px;
            border-radius: 50%;
            background: linear-gradient(to right, var(--primary-accent), var(--primary-hover));
            color: white;
            display: flex;
            align-items: center;
            justify-content: center;
            font-weight: 700;
            font-size: 0.92rem;
        }

        .step-content strong {
            display: block;
            margin-bottom: 6px;
            color: var(--text-main);
        }

        .step-content span {
            color: var(--text-muted);
            line-height: 1.7;
            font-size: 0.92rem;
        }

        /* DISCLAIMER */
        .disclaimer {
            margin-top: 35px;
            background: rgba(245, 158, 11, 0.08);
            border: 1px solid rgba(245, 158, 11, 0.3);
            color: #d97706;
            padding: 22px;
            border-radius: 22px;
            line-height: 1.8;
            font-size: 0.93rem;
        }

        /* BUTTONS */
        .btn-group {
            display: flex;
            gap: 20px;
            justify-content: center;
            margin-top: 42px;
            flex-wrap: wrap;
        }

        .btn {
            text-decoration: none;
            padding: 16px 30px;
            border-radius: 16px;
            font-weight: 700;
            transition: all 0.25s ease;
            min-width: 220px;
            text-align: center;
        }

        .btn-primary {
            background: linear-gradient(to right, var(--primary-accent), var(--primary-hover));
            color: white;
            box-shadow: 0 12px 25px rgba(37, 99, 235, 0.2);
        }

        .btn-primary:hover {
            transform: translateY(-2px);
            box-shadow: 0 18px 30px rgba(37, 99, 235, 0.24);
        }

        .btn-secondary {
            background: var(--card-bg);
            border: 1px solid var(--border-color);
            color: var(--text-main);
        }

        .btn-secondary:hover {
            transform: translateY(-2px);
            background: var(--bg-color);
        }

        /* FOOTER */
        .footer {
            text-align: center;
            margin-top: 45px;
            color: var(--text-muted);
            font-size: 0.9rem;
        }

        /* RESPONSIVE */
        @media (max-width: 900px) {
            .documentation-grid {
                grid-template-columns: 1fr;
            }
            .sticky-column {
                position: static;
            }
            .hero-section h1 {
                font-size: 2.3rem;
            }
        }

        @media (max-width: 600px) {
            .card {
                padding: 26px;
            }
            .btn {
                width: 100%;
            }
            .hero-section h1 {
                font-size: 2rem;
            }
        }
    </style>
</head>
<body>

    <div class="container">

        <!-- Theme Toggle Button -->
        <button class="theme-toggle-btn" id="themeToggle" onclick="toggleTheme()" type="button">
            <i class="fa-solid fa-moon"></i> <span id="themeToggleText">Dark Mode</span>
        </button>

        <!-- HERO -->
        <div class="hero-section">
            <div class="hero-badge">
                Model Documentation & Architecture
            </div>
            <h1>
                Understanding the Prediction Pipeline
            </h1>
            <p>
                This documentation explains the machine learning architecture, preprocessing pipeline, optimization strategy, and the reasoning behind simplifying telecom features for both predictive performance and end-user usability.
            </p>
        </div>

        <!-- MAIN GRID -->
        <div class="documentation-grid">
            
            <!-- LEFT SIDE -->
            <div class="card large-card">
                <div class="card-title">
                    <i class="fa-solid fa-brain"></i> Why XGBoost Classifier?
                </div>
                <p>
                    The prediction system uses the <strong>XGBoost (Extreme Gradient Boosting) Classifier</strong> because the target task involves predicting a binary customer state (Churn vs. No Churn) using a combination of numerical billing patterns (charges, tenure) and categorical features (contract type, payment method).
                </p>
                <p>
                    XGBoost was selected because it represents the state-of-the-art in gradient boosted decision trees for structured tabular datasets, providing high training efficiency, built-in regularization, and superior classification capabilities.
                </p>

                <div class="feature-list">
                    <div class="feature-item">
                        <strong>Regularized Gradient Boosting</strong>
                        <span>
                            XGBoost incorporates L1 (Lasso) and L2 (Ridge) regularization constraints to control tree complexity and prevent overfitting during split evaluations.
                        </span>
                    </div>

                    <div class="feature-item">
                        <strong>Imbalance Mitigation</strong>
                        <span>
                            Configures dynamic target weighting via the `scale_pos_weight` hyperparameter, which balances training weights based on the positive-to-negative sample ratio to handle class skewness.
                        </span>
                    </div>

                    <div class="feature-item">
                        <strong>Non-linear Separation</strong>
                        <span>
                            Enables decision split thresholds that naturally map complex, non-linear interactions between service tenures, product counts, and monthly rates.
                        </span>
                    </div>
                </div>

                <div class="highlight-box">
                    <div class="highlight-title">Why Not Linear / Basic Classifiers?</div>
                    <p>
                        Traditional linear models assume independent, linear interactions. Customer churn data containing high multi-collinearity and multi-service usage thresholds (where churn peaks at low tenure and moderate charges) is better modeled by decision-tree systems. XGBoost significantly outperformed traditional baseline classifiers during cross-validation.
                    </p>
                </div>
            </div>

            <!-- RIGHT SIDE -->
            <div class="sticky-column">
                
                <!-- ONE-HOT ENCODING -->
                <div class="card small-card">
                    <div class="card-title">
                        <i class="fa-solid fa-cubes"></i> One-Hot Encoding
                    </div>
                    <p>
                        One-Hot Encoding (OHE) was utilized to transform raw categorical columns (such as Contract, PaperlessBilling, and PaymentMethod) into numeric formats.
                    </p>
                    <div class="feature-list">
                        <div class="feature-item">
                            <strong>Dummy Variable Trap Avoidance</strong>
                            <span>
                                Configures `drop='first'` to drop the baseline column for each category, preventing collinearity issues in parameter calculations.
                            </span>
                        </div>
                        <div class="feature-item">
                            <strong>Algorithmic Compatibility</strong>
                            <span>
                                Ensures raw strings are converted into mathematical array formats required by gradient boosted tree algorithms.
                            </span>
                        </div>
                    </div>
                </div>

                <!-- GRID SEARCH -->
                <div class="card small-card">
                    <div class="card-title">
                        <i class="fa-solid fa-magnifying-glass-chart"></i> GridSearchCV
                    </div>
                    <p>
                        GridSearchCV was deployed to run exhaustive hyperparameter tuning over cross-validation folds, identifying optimal parameters for tree depth, estimators, and learning rates.
                    </p>
                    <div class="feature-list">
                        <div class="feature-item">
                            <strong>Exhaustive Optimization</strong>
                            <span>
                                Evaluates every parameter configuration across a defined parameter grid to prevent manual tuning bias.
                            </span>
                        </div>
                        <div class="feature-item">
                            <strong>Maximized ROC AUC</strong>
                            <span>
                                Optimizes model evaluation based on the Area Under the ROC Curve, balancing true positive and false positive rates.
                            </span>
                        </div>
                    </div>
                </div>

            </div>

        </div>

        <!-- FEATURE ENGINEERING -->
        <div class="card pipeline-card">
            <div class="card-title">
                <i class="fa-solid fa-code-fork"></i> Simplifying Features for the Model and the End User
            </div>
            <p>
                One important design decision in this project was preprocessing the raw customer usage and service variables into structured, derived feature groups to improve learning quality and simplify client-side entry.
            </p>
            <p>
                Tabular telecom data often contains detailed, correlated medical and account records. Direct usage of raw data fields can increase dimensionality, cause noise, and complicate user interaction.
            </p>

            <div class="pipeline">
                <div class="pipeline-step">
                    <div class="step-number">1</div>
                    <div class="step-content">
                        <strong>Service Add-On Mapping</strong>
                        <span>
                            Individual premium add-ons (such as OnlineSecurity, OnlineBackup, DeviceProtection, TechSupport, StreamingTV, and StreamingMovies) are mapped to numeric indicators (-1 for No, 0 for No Internet, and 1 for Yes) to construct consistent ordinal inputs.
                        </span>
                    </div>
                </div>

                <div class="pipeline-step">
                    <div class="step-number">2</div>
                    <div class="step-content">
                        <strong>Automatic Billing Consolidation</strong>
                        <span>
                            Aggregates specific payment channels containing "(automatic)" to compile a single, binary `is_automatic` feature, capturing customer financial convenience.
                        </span>
                    </div>
                </div>

                <div class="pipeline-step">
                    <div class="step-number">3</div>
                    <div class="step-content">
                        <strong>Standardization & Scaling</strong>
                        <span>
                            Applies `StandardScaler` to `MonthlyCharges` and `TotalCharges` dynamically on the server using training population stats to prevent scale bias from dominating prediction thresholds.
                        </span>
                    </div>
                </div>

                <div class="pipeline-step">
                    <div class="step-number">4</div>
                    <div class="step-content">
                        <strong>Ecosystem Integration Metrics</strong>
                        <span>
                            Derives `Product_Count` and flags `Is_High_Risk_Integration` (1 to 3 products) and `Is_Fully_Integrated` (5 or 6 products) variables to model the protective effect of customer service bundling on user loyalty.
                        </span>
                    </div>
                </div>
            </div>

            <div class="highlight-box">
                <div class="highlight-title">Design Philosophy</div>
                <p>
                    The objective was not only maximizing model accuracy and ROC AUC scores, but also creating a customer-centric analytics system that remains understandable, structured, and highly interactive for real-world business decisions.
                </p>
            </div>
        </div>

        <!-- DISCLAIMER -->
        <div class="disclaimer">
            <strong>Analytical Disclaimer:</strong>
            This platform is intended for analytical, educational, and demonstration purposes only. Predictions generated by the system should be interpreted as probability weights and not direct statements of customer action.
        </div>

        <!-- BUTTONS -->
        <div class="btn-group">
            <a href="{{ url_for('predict_model') }}" class="btn btn-primary">
                <i class="fa-solid fa-gauge-high"></i> &nbsp; Open Prediction Tool
            </a>
            <a href="/" class="btn btn-secondary">
                <i class="fa-solid fa-house"></i> &nbsp; Return to Home
            </a>
        </div>

        <!-- FOOTER -->
        <div class="footer">
            Customer Churn Predictor • XGBoost Classifier • One-Hot Encoding • GridSearchCV • Feature Engineering
        </div>

    </div>

    <script>
        document.addEventListener('DOMContentLoaded', () => {
            const savedTheme = localStorage.getItem('theme') || 'dark'; // Default to dark!
            document.documentElement.setAttribute('data-theme', savedTheme);
            updateThemeUI(savedTheme);
        });

        function toggleTheme() {
            const currentTheme = document.documentElement.getAttribute('data-theme');
            const newTheme = currentTheme === 'dark' ? 'light' : 'dark';
            document.documentElement.setAttribute('data-theme', newTheme);
            localStorage.setItem('theme', newTheme);
            updateThemeUI(newTheme);
        }

        function updateThemeUI(theme) {
            const btnText = document.getElementById('themeToggleText');
            const btnIcon = document.querySelector('.theme-toggle-btn i');
            if (theme === 'dark') {
                btnText.textContent = 'Light Mode';
                btnIcon.className = 'fa-solid fa-sun';
            } else {
                btnText.textContent = 'Dark Mode';
                btnIcon.className = 'fa-solid fa-moon';
            }
        }
    </script>
</body>
</html>