This project uses an LSTM neural network to predict air quality (PM2.5) from synthetic time-series data. It includes data generation, normalization, model training, and prediction visualization. The results demonstrate how deep learning can forecast pollution levels

Deep learning air-pollution forecaster with Pytorch

Analysis of Air Pollution prediction and time-series forecast of PM2.5 Pollutant using Machine Learning Algorithms (SVM, Decision Tree and Random Forest) and Deep Learning Algorithms (CNN, Bi-LSTM). Also considered for improved performances is random search hyper-parameter tuning using Ray-Tune with HyperBand Scheduler strategy.

A comprehensive time series forecasting project comparing statistical, machine learning, and deep learning models for air pollution prediction. Features data exploration, model implementation (ARIMA, SARIMAX, Prophet, Random Forest, XGBoost, LightGBM, LSTM, DeepAR), and performance evaluation.

Advanced Machine Learning Framework for Forecasting Air Pollution in Delhi (2025-2030) using Hybrid Deep Learning Architecture

This repository presents an Air Quality Prediction system using IoT sensors and Deep Learning. It collects real-time environmental data (PM2.5, CO₂, temperature, humidity) via IoT devices and applies LSTM-based models for accurate forecasting, enabling smart city planning and pollution control.

This is a repository that contains the codes and the some of the processed data used for the undergraduate thesis titled "Forecasting Air Pollution of Dhaka City with Meteorological Parameters Using Deep Learning-Based Approach". For more details please contact me at: rvnrafsan@gmail.com

Using Deep learning methods for Time Series Forecasting of Air Pollution

Air pollution forecasting of O3 and NO2 using machine learning and deep learning

LSTM-based deep learning model to forecast air pollution (PM2.5) using time-series environmental data.

AirAware – Air Quality Predictor using Deep Learning (LSTM) A deep learning project that predicts air quality levels using historical pollution data. The model uses an LSTM neural network implemented in Python to forecast air quality indicators and help monitor environmental conditions.

A deep learning solution for time series forecasting of PM2.5 air pollution concentrations in Beijing using bidirectional LSTM networks with attention mechanisms and ensemble learning.

A deep‑learning powered time‑series model that accurately forecasts hourly CO concentrations to help urban planners and communities monitor and mitigate air pollution in real time.

The AQI Prediction and Forecasting System using ML and DL analyzes historical air pollution data to predict future Air Quality Index values. Machine Learning and Deep Learning models are used to improve prediction accuracy and identify pollution trends. This helps in environmental monitoring and protecting public health.

A deep learning project using LSTM networks to predict air quality trends based on historical pollution data. Designed to spot patterns in time-series data and forecast future air quality levels with style and accuracy.

Air Quality Index (AQI) Prediction Using Deep Learning with Seasonal Decomposition is a data-driven project that forecasts air pollution levels using deep learning techniques. It analyzes historical environmental data and seasonal patterns to improve prediction accuracy, helping support better environmental monitoring and decision-making. 🌍

Air Quality Prediction using Multivariate Time-Series Forecasting. This project uses machine learning and deep learning models to analyze historical air pollution data and predict future AQI values based on multiple environmental variables such as PM2.5, PM10, NO₂, CO, and weather conditions.

An Air Quality Index (AQI) prediction project uses machine learning or deep learning techniques to forecast future air pollution levels based on historical data, pollutant concentrations, and meteorological factors. The goal is to provide timely, actionable information for public health and environmental management.

The AQI Prediction project uses ML to forecast air pollution based on environmental data. It leverages pollutant levels (PM2.5, PM10, NO2, CO, SO2, O3), temperature, humidity, and wind speed to predict AQI. Using linear regression and deep learning, it aids proactive air quality management and pollution control.

LSTM-based deep learning model to forecast PM2.5 air pollution using multivariate time series data. The project includes data preprocessing, feature encoding, scaling, sequence generation, model training, and evaluation on real-world environmental datasets

Worked on diffrent projects 1. Mobile price segmentation 2. Walmart Market Data Visualization & Prediction 3. Bank fraud transaction detection 4. Chest X-Ray: Predict Pneumonia 5. Comparing Deep Learning Optimizers 6. China Air pollution forecasting

Air Quality Prediction System uses Python, Deep Learning, YOLOv8, LSTM, TensorFlow, and OpenCV to analyze traffic videos and predict AQI, moisture, and pollution levels in real time. It detects vehicles and pollutants, forecasts trends, and provides smart-city environmental insights with accurate, data-driven air quality alerts.

A Hybrid CNN-GRU deep learning model for early prediction of respiratory diseases caused by air pollution. Combines CNN for feature extraction and GRU for temporal patterns, enhanced with self-attention. Achieves high accuracy and supports real-time pollutant-based health risk forecasting.

Air pollution is one of the most concerning problems world- wide. It leads to the necessary time series forecasting particulate mat- ter (PM) concentrations. In this study, we propose an effective way to improve the results of PM10 and PM25 forecasting tasks by combin- ing meteorological features and timestamp information using datasets from Hanoi, Vietnam. By conducting five deep learning models, MLP, 1D-CNN, LSTM, Bi-LSTM, and Stacked LSTM, the experimental re- sults show that adding more timestamps information helped improve the results in 68% of cases compared to other methods. Furthermore, the Vanilla LSTM model with combined features will give better results in the long-term forecast than the other. We plan to extend our ap- proach by continuing to combine more features from multiple sources and applying more robust models for the main problem.