codebenders-datathon/ML_PIPELINE_REPORT.txt at main · devcolor/codebenders-datathon · GitHub

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

BISHOP STATE ML PIPELINE - SUMMARY REPORT
================================================================================
Generated: 2026-02-21 12:59:23

DATASET OVERVIEW
--------------------------------------------------------------------------------
Total Students: 4,000
Total Course Records: 4,000

MODEL PERFORMANCE SUMMARY
--------------------------------------------------------------------------------

1. RETENTION PREDICTION MODEL
   Algorithm: XGBoost Classifier
   Features Used: 23
   Test Set Performance:
     - Accuracy: 0.7238
     - AUC-ROC: 0.6134

   Risk Distribution:
     Critical Risk             0 (  0.0%)
     High Risk                82 (  2.1%)
     Moderate Risk         2,195 ( 54.9%)
     Low Risk              1,723 ( 43.1%)

2. EARLY WARNING SYSTEM
   Algorithm: Composite Risk Score (Retention + Performance Metrics)
   Approach: Aligned with retention predictions to eliminate contradictions
   Alert Distribution:
     URGENT          0 (  0.0%)
     HIGH           21 (  0.5%)
     MODERATE    2,210 ( 55.2%)
     LOW         1,769 ( 44.2%)

3. TIME TO CREDENTIAL PREDICTION
   Algorithm: XGBoost Regressor
   Mean Predicted Time: 2.97 years
   Median Predicted Time: 2.96 years

4. CREDENTIAL TYPE PREDICTION
   Algorithm: Random Forest Classifier
   Predicted Distribution:
     No Credential         4,000 (100.0%)

5. GATEWAY MATH SUCCESS PREDICTION (NEW!)
   Algorithm: XGBoost Classifier
   Students with Gateway Math Data: 4,000
   Average Pass Probability: 0.0%

   Gateway Math Risk Distribution:
     High Risk             4,000 (100.0%)

6. GATEWAY ENGLISH SUCCESS PREDICTION (NEW!)
   Algorithm: XGBoost Classifier
   Students with Gateway English Data: 4,000
   Average Pass Probability: 0.0%

   Gateway English Risk Distribution:
     High Risk             4,000 (100.0%)

7. FIRST-SEMESTER LOW GPA (<2.0) PREDICTION (NEW!)
   Algorithm: XGBoost Classifier
   Average Low GPA Probability: 13.1%
   Students Predicted Low GPA: 231

   Academic Risk Level Distribution:
     Low Risk              3,078 ( 77.0%)
     Moderate Risk           597 ( 14.9%)
     High Risk               258 (  6.5%)
     Critical Risk            67 (  1.7%)

OUTPUT: DATABASE TABLES
--------------------------------------------------------------------------------
1. student_predictions (Table)
   - Student-level data with all predictions
   - 4,000 students
   - 164 columns

2. course_predictions (Table)
   - Course-level data with predictions
   - 4,000 records
   - 159 columns

3. ml_model_performance (Table)
   - Model performance metrics
   - Training timestamps and notes

PREDICTION COLUMNS ADDED
--------------------------------------------------------------------------------
Retention:
  - retention_probability (0-1 scale)
  - retention_prediction (0=Not Retained, 1=Retained)
  - retention_risk_category (Critical/High/Moderate/Low Risk)

Early Warning:
  - at_risk_probability (0-1 scale)
  - at_risk_prediction (0=Not At Risk, 1=At Risk)
  - at_risk_alert (URGENT/HIGH/MODERATE/LOW)
  - risk_score (0-100 comprehensive risk score)

Time to Credential:
  - predicted_time_to_credential (years)
  - predicted_graduation_year (year)

Credential Type:
  - predicted_credential_type (0-3 numeric)
  - predicted_credential_label (text label)
  - prob_no_credential, prob_certificate, prob_associate, prob_bachelor

Gateway Math Success:
  - gateway_math_probability (0-1 scale)
  - gateway_math_prediction (0=Won't Pass, 1=Will Pass)
  - gateway_math_risk (High Risk/Moderate Risk/Likely Pass/Very Likely Pass)

Gateway English Success:
  - gateway_english_probability (0-1 scale)
  - gateway_english_prediction (0=Won't Pass, 1=Will Pass)
  - gateway_english_risk (High Risk/Moderate Risk/Likely Pass/Very Likely Pass)

First-Semester GPA < 2.0 Risk:
  - low_gpa_probability (0-1 scale)
  - low_gpa_prediction (0=Adequate GPA, 1=Low GPA)
  - academic_risk_level (Low Risk/Moderate Risk/High Risk/Critical Risk)

================================================================================
PIPELINE COMPLETE!
================================================================================