LookBeforeYouLeap

Leveraging predictive ML models to improve Automotive Safety and Travel Time through Twitter Sentiment Analysis and Traffic Record Analysis

Authors: Hannah Do, Kamil Sachryn

Project Summary

Retrieval of tweets through Snscape, Tweepy, and geocoding through ArcGIS - 1
Text processing through various NLP processing and Vader methods (sentiment analysis) - 1
Collection of accident record and feature selection - 2
Integration of twitter features into traffic accident records - merging the two datasets based on the distance between instances - 3
SMOTE and random instance generation for balancing the class - 3
ML predictions and evaluation (processed files are in 'merged' folder) - 4

Name		Name	Last commit message	Last commit date
Latest commit History 95 Commits
.ipynb_checkpoints		.ipynb_checkpoints
02032022		02032022
data_traffic		data_traffic
data_twitter		data_twitter
figures		figures
locat		locat
merged		merged
presentation_slides		presentation_slides
.DS_Store		.DS_Store
1_Getting_Twitter_Data.ipynb		1_Getting_Twitter_Data.ipynb
2_Getting_Traffic_Data.ipynb		2_Getting_Traffic_Data.ipynb
3_Integrating_Dataframes_and_Random_Sampling.ipynb		3_Integrating_Dataframes_and_Random_Sampling.ipynb
4_ML_Prediction.ipynb		4_ML_Prediction.ipynb
Final Paper ML.pdf		Final Paper ML.pdf
ProjectProposalML.pdf		ProjectProposalML.pdf
README.md		README.md
final presentation ML.pdf		final presentation ML.pdf
poster_LookBeforeYouLeap.pdf		poster_LookBeforeYouLeap.pdf
poster_image.png		poster_image.png