A machine-learning project
btp-berlin is a experimental learning project in which we trying to train a model to be able to predict bike traffic in certain areas depending on weather data and also possible unpredictable events like covid and the lockdows. btp-berlin is trained in data from 2015 to 2022.
For Documentation please use the Notebooks
Data preperation script:
- as a Notebook :
ml_reg_training.ipynb - only code :
ml_data_prep.ipynb
Training the Model:
- as a Notebook :
ml_reg_training.py - only code :
ml_data_prep.py
Raw source data:
- wind.csv
- air_temp.csv
- sun.csv
- precipation.csv
- airpressure.csv
- radverkehr_final.csv
Merged single file output from 'ml_data_prep`
- clean_df.csv