Hybrid Neural Networks and Boosted Regression Tree Models for Predicting Roadside Particulate Matter

Abstract

This paper examines the application of artificial neural network (ANN) and boosted regression tree (BRT) methods in air quality modelling. The methods were applied to developing air quality models for predicting roadside particle mass concentration (PM 10 , PM 2.5 ) and particle number counts (PNC) based on air pollution, traffic and meteorological data from Marylebone Road in London. Elastic net, Lasso and principal components analysis were used as feature selection methods for the ANN models to reduce the number of predictor variables and improve their generalisation. The performance of the ANN with feature selection (ANN hybrid) and the BRT models was evaluated and compared using statistical performance metrics. The performance parameters include root mean square error (RMSE), fraction of prediction within a factor of two of the observation (FAC2), mean bias (MB), mean gross error (MGE), the coefficient of correlation (R) and coefficient of efficiency (CoE) values. The input variables selected by the elastic net produced the best performing ANN models. The ANN hybrid produced models performed only slightly better than the BRT models. The R values of the ANN elastic net and BRT models were 0.96 and 0.95 for PM 10 , 0.96 and 0.96 for PM 2.5 and 0.89 and 0.87 for PNC, respectively. Their corresponding CoE values were 0.72 and 0.70 for PM 10 , 0.74 and 0.76 for PM 2.5 and 0.81 and 0.71 for PNC respectively. About 80-99% of all the model predictions are within a factor of two of the observed particle concentrations. The BRT models offer more advantages regarding model interpretation and permit feature selection. Therefore, the study recommends the use of BRT over ANN where the model interpretation is a priority.

Shallow Landslide Susceptibility Modeling Using the Data Mining Models Artificial Neural Network and Boosted Tree

“…In particular, the BT model had a higher accuracy (about 8%) than the ANN model. In other fields (using both models) [53,54], BT reported a better performance than ANN. The results of this study demonstrate the benefits of selecting optimal data mining techniques in landslide susceptibility modeling.…”

The used of the Boosted Regression Tree Optimization Technique to Analyse an Air Pollution data.

“…The aim of the second step of the experiment was to evaluate the predictive performance of the different models by varying the interaction depth parameter by using different k-fold cross-validations, where the value of k was set as 10 during model fitting. Basically, a spatial version of the k-fold cross-validation method used by [15] was used and implemented by using an R package and the gbm library [25]. Briefly, the rationale for adopting this approach is that the cross-validation method has the advantage of using all the data for both training and validation by repeating the process k-times on different combinations of subsamples and calculating the mean performance of the v-models.…”

“…Briefly, the rationale for adopting this approach is that the cross-validation method has the advantage of using all the data for both training and validation by repeating the process k-times on different combinations of subsamples and calculating the mean performance of the v-models. The k-fold cross-validation method partitions the data into k-subsets, where k-models are built based on the 'k minus 1' subsets and the model performance is tested based on the last remaining subset [25]; this process is repeated several times with different number of trees for model parameter until the optimum number of trees with the minimal error is determined.…”

“…All the models were fitted in the R package R 3.3.2 using the 'gbm' or 'generalized boosted machine' package version 2.1 [27], [25]. Each model was trained within 1 hour by using a modern computer with an Intel Core i7 (3632QM) CPU.…”

“…The final models were tested to the holdout test fraction with the optimum number of trees and were then evaluated by using error bias analyses to compare the observed PM 10 concentration values with the predicted values according to a range of evaluation statistics namely the factor of two (FAC2), Root Mean Square Error (RMSE), correlation coefficient (R), coefficient of efficiency (COE) and index of agreement (IOA) [25], [29].…”