https://docs.google.com/document/d/1qwZhqqooVG5XMvkUdVPrp-sd_1HF5tTXXNeJ8xwqmuI/edit?usp=sharing
https://docs.google.com/presentation/d/12Cq33mqEqm6_ZKWGDrCZi1hkaJzUIzgTrgtq8ba6G04/edit?usp=sharing
For maps of stations: http://www.weather.gov.sg/climate-historical-daily
Statistics of population:
https://www.singstat.gov.sg/find-data/search-by-theme/population/geographic-distribution/latest-data
Conventions for model training and documentation: https://docs.google.com/document/d/14CumPA9qrzm4UgbWdu4Z207IkrD-B1KYKaRphHI6RTM/edit?usp=sharing
NOTES: Ignore semakau data and sentosa island data.
Possible useful links:
https://github.com/ngbolin/DengAI
Chinese paper on year window dengue prediction:
https://www.biorxiv.org/content/biorxiv/early/2019/09/06/760702.full.pdf
Google groups: https://groups.google.com/forum/#!forum/iot-datathon-30-aisavelives---dengue-outbreak-forecasting
Dengue NEA site (but no past data): https://www.nea.gov.sg/dengue-zika/dengue/dengue-clusters
Documents Explanation:
datathon3-data/rainfall/_combined$PLACE$.csv: aggregating ranifall to temperature stations
datathon3-data/rainfall/_zeroes$PLACE.csv: just like combine place, but replacing unknown parts with 0.
datathon3-data/daily_tabulated_data.csv: combining the rainfall and temperature, daily
datathon3-data/weekly_tabulated_data.csv: averaging over temperature, and rafall_zeroes
weekly_labeled_data_deleted: removed temperature stations and rainfall stations: sentosa, pulau ubin, semakau, jurong-island khatib and tuas south: remove population (since they are zeros), but keep rainfall and temperature stations