Comparative Analysis of Machine Learning Techniques for Predicting Air Quality in Smart Cities

Abstract

Dealing with air pollution presents a major environmental challenge in smart city environments. Real-time monitoring of pollution data enables local authorities to analyze the current traffic situation of the city and make decisions accordingly. Deployment of the Internet of Things-based sensors has considerably changed the dynamics of predicting air quality. Existing research has used different machine learning tools for pollution prediction; however, comparative analysis of these techniques is required to have a better understanding of their processing time for multiple datasets. In this paper, we have performed pollution prediction using four advanced regression techniques and present a comparative study to determine the best model for accurately predicting air quality with reference to data size and processing time. We have conducted experiments using Apache Spark and performed pollution estimation using multiple datasets. The Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) have been used as evaluation criteria for the comparison of these regression models. Furthermore, the processing time of each technique through standalone learning and through fitting the hyperparameter tuning on Apache Spark has also been calculated to find the best-fit model in terms of processing time and lowest error rate.INDEX TERMS IoT, smart city, air quality index (AQI), data mining, Apache Spark.

Air Quality Prediction in Smart Cities Using Machine Learning Technologies Based on Sensor Data: A Review

“…N/S [14] AQ, MET YES To obtain a sequence pattern. N/S [39] AQ, MET NO RFR reduces overfitting, detects peak values.…”

“…Comparative Analysis of Machine Learning Techniques for Predicting Air Quality in Smart Cities [39]: Ameer et al used different models for predicting air quality, such as DTR, Random Forest Regression (RFR), MLP and GBR. The dataset used in this study included year, month, day, hour, season, PM 2.5 , dew point, temperature, humidity, pressure, combined wind direction, accumulated wind speed, hourly precipitation, accumulated precipitation.…”

“…Tao et al [14] also mentioned the superiority of LSTM compared to RNN and confirmed the advantage of the bidirectional model. According to Ameer et al [39] compared to a decision tree, gradient boosting and multilinear perception, random forest obtained better results by reducing overfitting and detecting peak values. Although, on the other hand, Li and Ngan [61] mentioned that random forest could have a challenge with fitting a wide variety of data distribution.…”

PM2.5 Prediction using Machine Learning Hybrid Model for Smart Health

“…At present, many machine learning techniques have been proposed for solving air pollution prediction problems based on simple regression models.Authors in [3] have performed to estimate PM2.5 using random forest model and two other traditional regression models, the random forest shows the high accuracy in predicting the PM2.5 concentration. In [1] authors made a comparative study of machine learning techniques to predict the quality of air using Apache spark with multiple data sets and concluded that the random forest was a best technique in prediction but it actually work well for small size dataset and performs well only on classification problems. In [6] author proposed new model based on LSTM(Long Short-Term Memory) to forecasting PM2.5 based on the historical data and in [18] author achieves in predicting PM2.5,NO2,SO2 air pollutants concentrations with ARIMA(Auto Regressive Integrated Moving Average, Simple and Exponential Weighted Moving Average , KF algorithm and obtained the optimal result but the data handling and processing time was not discussed.…”

Resource Usage and Performance Trade-offs for Machine Learning Models in Smart Environments

“…Internet of Things (IoT) applications are processing growing amounts of sensor information to monitor everyday environments. Typical application examples include those supporting the elderly in smart homes [1], the ones monitoring environmental parameters in smart cities [2], smart health applications on wearable devices [3] and fault detection solutions for smart manufacturing [4].…”