BGP_2021_B

Big Geodata Processing -2021 - Group B

Group project with the goal to predict the observer intensity (defined as the number of observers per 1km^2-block and per hour/per week) in time or in space, based n data from the Citizen Science project waarneming.nl.

The data was provided through the following data base:
host: gip.itc.utwente.nl
port: 5434
database: c211
user: sxxxxxx
password: xxxxxx\

Instructions

1. Folder 1_data_preparation: The three notebooks daily_observer_intensity, weekly_observer_intensity and monthly_observer_intensity prepare the original datasets. Each one of the notebooks creates a new single table with all the fields required for further processing in the indicated temporal resolution.
2. Folder 2_data_exploration: The notebook expl_data_statistics performs some exploratory analysis on the data from the tables creating in step 1). The notebook Map creates a map showing the spatial distribution of observer intensity in the Netherlands.
3. Folder 3_modeling: The notebook ML models the observer intensity using a supervised machine learning algorithm (random forest regression) in several configurations, using the data created in step 1).