GitHub - amartyads/accidents-as-anomalies: Tries to predict whether an accident is fatal or not given environmental factors, using anomaly detection techniques. Work done on the kaggle dataset available at https://www.kaggle.com/daveianhickey/2000-16-traffic-flow-england-scotland-wales

amartyads / accidents-as-anomalies Public

Notifications You must be signed in to change notification settings
Fork 0
Star 0

Tries to predict whether an accident is fatal or not given environmental factors, using anomaly detection techniques. Work done on the kaggle dataset available at https://www.kaggle.com/daveianhickey/2000-16-traffic-flow-england-scotland-wales

0 stars 0 forks Branches Tags Activity

Star

Notifications

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 1 Commit
dataset		dataset
datavis graphs		datavis graphs
.gitignore		.gitignore
01datasetvis.py		01datasetvis.py
02gru.py		02gru.py
03autoencanomaly.py		03autoencanomaly.py
04randomforest.py		04randomforest.py
05grupredict.py		05grupredict.py
06gruparampredict.py		06gruparampredict.py
07parampredict.py		07parampredict.py
08ocsvm.py		08ocsvm.py
09gan.py		09gan.py
10dashapp.py		10dashapp.py
11dashrealtrial.py		11dashrealtrial.py
12dashgui.py		12dashgui.py
13dashfinal.py		13dashfinal.py
README.txt		README.txt
autoencoder.h5		autoencoder.h5
datavis report.pdf		datavis report.pdf
datavisobs.txt		datavisobs.txt
experiment report.pdf		experiment report.pdf
randomforesth2o.py		randomforesth2o.py
test2png.png		test2png.png
trafficvis.html		trafficvis.html

Repository files navigation

This code was written to work with the dataset available at https://www.kaggle.com/daveianhickey/2000-16-traffic-flow-england-scotland-wales

The work was done using the Spyder editor for Python, hence the code is broken into cells which are individually runnable.

Details of the work can be found in 'datavis report.pdf' and 'experiment report.pdf'. They can also serve as an introduction for the datasets.

Raw latex files are in the datavis graphs folder.

The code includes temporary programs and test code, left here for completion's sake.

Please download the dataset from the aforementioned link, and paste into the dataset folder.