Weather_Prediction is a set of AI models predicting temperature based on a number of factors given in the input file.
Examplary data included in this project was borrowed from imgw.pl
Project currently contains linear and neural network models:
- pres_linear_models.ipynb contains linear models
- pres_neural_network.ipynb contains neural network model
- Linear Regression
- Linear Auto Regressive
- Linear Auto Regressive with lag
- Feedforward artificial neural networks
Data from cities:
- GDANSK
- WARSZAWA
- BIALYSTOK
- BIELSKO-BIALA
- SZCZECIN
- WROCLAW
Features used:
- Minimum Temperature
- Maximum Temperature
- Average Temperature
- Sum Of Falls
- Kind Of Falls
- Cloudiness
- Wind Speed
- Humidity
- Pressure
- In order to imitate our environment, we recomend installing Visual Studio Code with Jupyter extension. Make sure you have the required libraries installed.
- Download raw data from imgp.pl.
- Unzip files to folders in data as you wish(in our example, we sorted individual .csv files into folders describing years). Use csv_converter.ipynb in combination with getCSVFiles functions to merge them together into one resulting file result.csv.
- Open one of files containing presentations of AI models and run the program with Jupyter.
- Notice the content of file change after running the program. Graphs changed according to your input files. Voilà!
- statsmodels
- matplotlib
- pandas
- numpy
- csv
- os