HackMDStewards Presentation (R9)
EDA stats
Semi Supervised Pipeline
Heuristic labels
IP address cluster > 2 AND GitHub account creation date > '2021-03-09'
Sensitivity vs Specificity
Sensitivity vs Specificity (1)
Sensitivity vs Specificity (2)
Heuristic EDA
General stats
Recent user share: 43.0%
IP cluster
User count: 1785
User share: 16.1%
Contribution total: 24535
Contribution share: 14.5%
Amount total: 80415.05
Amount share: 4.5%
Recent user composition: 72.2%
Recent user share: 27.0%
Wallet cluster
User count: 788
User share: 7.1%
Contribution total: 17847
Contribution share: 10.6%
Amount total: 690602.19
Amount share: 38.6%
Recent user composition: 26.3%
Recent user share: 4.3%
Wallet AND IP cluster
User count: 127
User share: 1.1%
Contribution total: 2581
Contribution share: 1.5%
Amount total: 17417.29
Amount share: 1.0%
Recent user composition: 26.0%
Recent user share: 0.7%
Model Results
Model 1 (RF importance)
Model 1 (DT)
Model 2 (RF importance)
Stats after flagging
- R9 users: 11093
- Users created during R9: 43%
Summary stats (1)
Total Contributions (original): 195147
Total Contributions (modified): 195147
Matched Contributions (original): 176662
Matched Contributions (modified): 160026
Summary stats (2)
Total Contributors (original): 12906
Total Contributors (modified): 11420
Median Median USDT Amount (original): 1.66
Median Median USDT Amount (modified): 1.73
Median Contrib Count per User (original): 7.00
Median Contrib Count per User (modified): 7.00
Mean Count (original): 14.98
Mean Count (modified): 16.93
Quadratic Funding stats
Total Grant Funding (original): 500000.00
Total Grant Funding (modified): 500000.00
Mean Grant Funding (USDT, original): 841.75
Mean Grant Funding (USDT, modified): 836.12
Median Grant Funding (USDT, original): 1.16
Median Funding (USDT, modified): 1.14
Grant count (original): 594
Grant count (modified): 598
Max payout (USDT): 533014
