Open AccessImproving air quality management using gradient boosting based hierarchical temporal memory neural networks and fuzzy based classification based regression tree
Author(s) -
S Sagayaraj,
N Vetrivelan
Publication year - 2018
Publication title -
international journal of engineering and technology
Language(s) - English
Resource type - Journals
ISSN - 2227-524X
DOI - 10.14419/ijet.v7i2.9.9229
Subject(s) - air quality index , particulates , artificial neural network , wind speed , environmental science , mean squared error , decision tree , pollutant , boosting (machine learning) , nitrogen dioxide , normalization (sociology) , gradient boosting , air pollution , computer science , random forest , artificial intelligence , meteorology , statistics , mathematics , geography , chemistry , organic chemistry , sociology , anthropology
In recent years, air pollution introduces different biological molecules, particulate and several harmful materials which affect the human health and activities. So, the quality of the air should be maintained for avoiding the above issues. To manage the air quality initially the meteorological data have been collected from Ariyalur that includes the condition of air, data collected date, high and low temperature, wind speed, wind direction and relative humidity. The collected data has to be preprocessed by applying the normalization and data mining techniques and those preprocessed data’s are used to predict the pollutants and the concentration level of the pollutants such as sulfur dioxide (SO2), carbon monoxide (CO), nitrogen dioxide (NO2), and nitric oxide (NO). Then the particulate matter level in the air has to be predicted by Gradient Boosting based Hierarchical Temporal Memory Neural Networks (BHTMNN). From the predicted value the strength of the pollutants is classified by using the Fuzzy based Classification based Regression Tree (FCART) which is used to recognize the disease arises in the human respiratory system. Then the performance of the proposed system is evaluated using the mean square error, classification accuracy, sensitivity and specificity.