Final Mo'Mentum Slides

Final Mo'Mentum Slides

Transcript

1. Consulting Proposal September 4, 20XX Mo’Mentum Optimize your mo’mentum for social change Peter Winslow
4. Many resources for sharing petitions!
5. Many resources for sharing petitions!
6. Victory! Ok, let’s do this! Many resources for sharing petitions!
7. Is my petition good enough? Many resources for sharing petitions! Little help for writing petitions...
8. Just what I was looking for! Mo’Mentum Many resources for sharing petitions! Little help for writing petitions...
9. Just what I was looking for! Mo’Mentum My Petition Many resources for sharing petitions! Little help for writing petitions...
10. Just what I was looking for! Mo’Mentum Many resources for sharing petitions! Little help for writing petitions... My Petition Probability of success Time scale to reach signature goal
11. Data Collection Change.org sitemap Petition urls
12. Petition urls Petition id’s Data Collection Change.org sitemap
13. Petition urls Petition id’s Petition text and metadata Change.org API Data Collection Change.org sitemap
14. Over ~ 40,000 Petitions Data Collection Petition urls Petition id’s Petition text and metadata Change.org API Change.org sitemap
15. Feature Engineering Text data Stopwords, Lemmatization
16. Sentiment, POS Tagging, word/sentence counts, ... Text data Stopwords, Lemmatization Features Feature Engineering
17. Sentiment, POS Tagging, word/sentence counts, ... Text data Stopwords, Lemmatization Features Metadata Success/Failure Feature Engineering Signature Accumulation Rate
18. Metadata Success/Failure Signature Accumulation Rate Random Forest Classifier Sentiment, POS Tagging, word/sentence counts, ... Text data Stopwords, Lemmatization Features Feature Engineering
19. Accurate Prediction of Petition success Train-Test-evaluation split with 5-fold CV AUC = 87.4%
20. Metadata Success/Failure Signature Accumulation Rate Random Forest Classifier Sentiment, POS Tagging, word/sentence counts, ... Text data Stopwords, Lemmatization Features Feature Engineering Gradient Boosting Regressor
21. Accurately Predicts signature accumulation rate Train-Test-evaluation split with 5-fold CV RMSE/Y Max = 15% 81% within 1 92% within 2
22. Peter Winslow The Professional PhD + 1 Postdoc in theoretical High Energy Physics and Cosmology. Origin of matter in the Universe New Father! Kiana Winslow, born Nov. 29th 2016
23. Backup Slides
24. Backup Slides
25. Backup Slides
26. Algorithms: Classification Random Forest Classifier (Scikit-Learn) Predict success/failure of petition Reasons for choosing: • Lots of complication yet resistant to overfitting Challenges: • Class imbalance in the data Validation: Train-Test-evaluation split with 5-fold CV Backup Slides
27. Algorithms: Regression GradientBoostingRegressor (Scikit-Learn) Predict signature accumulation rate Reasons for choosing: • Many features, highly non-linear, can return predicted “quantiles” Challenges: • The right evaluation metric? Validation: Train-Test-evaluation split with 5-fold CV Backup Slides
28. Backup Slides Calculation of signature accumulation rate
29. Insights: Backup Slides New Year’s resolution boom and bust?