Timeseries forecasting of air pollution using LSTM Recurrent neural network

The study is conducted as as an end-of-course project for the "Machine Learning systems for Data Science" course of the Master Degree in Digital Humanities and Digital Knowledge of the University of Bologna, held by professors Stefano Lodi and Elisabetta Ronchieri.

The purpose of the project is to replicate the work process used for the case study in the research paper "Prediction of air quality in Jakarta during the COVID-19 outbreak using long short-term memory machine learning" by Wihayati and F W Wibowo 2021 IOP Conf. Ser.: Earth Environ. Sci. 704 012046.

Repository content

The repository contains the following folders:

• 'data' with data files in csv format;
• 'reference_paper' with the PDF containing the reference paper;
• 'scripts' with Jupyter Notebook files

Resources and tecchnologies used for the project

Datasets

The data are downloaded from https://data.jakarta.go.id/dataset/indeksstandar-pencemaran-udara-ispu-tahun-2020

Tools

As a tool for data manipulation, visualisation and analysis we chose Google Colab environment and various Python libraries, mainly:

• Pandas, in order to read .csv datasets and convert them to Python-friendly dataframes;
• Numpy;
• tensorflow.keras;
• matplotlib

Methods

• Data cleaning
• Data visualisation
• LSTM RNN

Contributors

• Anastasiya Sopyryaeva
• Manu Srivastava