Construction and demolition waste (DW) generation information has been recognized as a tool for providing useful information for waste management. Recently, numerous researchers have actively utilized artificial intelligence technology to establish accurate waste generation information. This study investigated the development of machine learning predictive models that can achieve predictive performance on small datasets composed of categorical variables. To this end, the random forest (RF) and gradient boosting machine (GBM) algorithms were adopted. To develop the models, 690 building datasets were established using data preprocessing and standardization. Hyperparameter tuning was performed to develop the RF and GBM models. The model performances were evaluated using the leave-one-out cross-validation technique. The study demonstrated that, for small datasets comprising mainly categorical variables, the bagging technique (RF) predictions were more stable and accurate than those of the boosting technique (GBM). However, GBM models demonstrated excellent predictive performance in some DW predictive models. Furthermore, the RF and GBM predictive models demonstrated significantly differing performance across different types of DW. Certain RF and GBM models demonstrated relatively low predictive performance. However, the remaining predictive models all demonstrated excellent predictive performance at R² values > 0.6, and R values > 0.8. Such differences are mainly because of the characteristics of features applied to model development; we expect the application of additional features to improve the performance of the predictive models. The 11 DW predictive models developed in this study will be useful for establishing detailed DW management strategies.

Recently, artificial intelligence (AI) technologies have been employed to predict construction and demolition (C&D) waste generation. However, most studies have used machine learning models with continuous data input variables, applying algorithms, such as artificial neural networks, adaptive neuro-fuzzy inference systems, support vector machines, linear regression analysis, decision trees, and genetic algorithms. Therefore, machine learning algorithms may not perform as well when applied to categorical data. This article uses machine learning algorithms to predict C&D waste generation from a dataset, as a way to improve the accuracy of waste management in C&D facilities. These datasets include categorical (e.g., region, building structure, building use, wall material, and roofing material), and continuous data (particularly, gloss floor area), and a random forest (RF) algorithm was used. Results indicate that RF is an adequate machine learning algorithm for a small dataset consisting of categorical data, and even with a small dataset, an adequate prediction model can be developed. Despite the small dataset, the predictive performance according to the demolition waste (DW) type was R (Pearson’s correlation coefficient) = 0.691–0.871, R² (coefficient of determination) = 0.554–0.800, showing stable prediction performance. High prediction performance was observed using three (for mortar), five (for other DW types), or six (for concrete) input variables. This study is significant because the proposed RF model can predict DW generation using a small amount of data. Additionally, it demonstrates the possibility of applying AI to multi-purpose DW management.

Introduction: Senior managers’ attitudes towards safety are very important regarding the safety practices in an organization. The study is to describe the current situation of senior managers′ attitudes towards safety in the Chinese coal industry. Method: We evaluated the changing trends as well as the reasons for these changes in the Chinese coal industry in 2009 and in 2014 with 168 senior manager samples from large Chinese state-owned coal enterprises. Evaluations of 15 safety concepts were performed by means of a questionnaire. Results and Conclusions: Results indicate that, in 2014, three concepts were at a very high level (mean > 4.5), and six were at a relatively high level (4.5 > mean > 4.0). Analyses of changing trends revealed that nine concepts improved significantly, while four greatly declined in 2014 compared to those in 2009. The data reported here suggest that the reasons for the significant improvement with respect to the nine concepts include the improvement in social and legal environments, the improvement of the culture of social safety, workers′ safety demands being met, and scientific and technical advances in the coal industry. The decline of the four concepts seemed to be caused by a poor awareness of managers in the coal industry that safety creates economic benefits, insufficient information on safety, inadequate attention to the development of a safety culture and safety management methods, and safety organizations and workers′ unions not playing their role effectively. Practical Applications: We therefore recommend strengthening the evidence that safety creates economic benefits, providing incentives for employees to encourage their participation in safety management, and paying more attention to the prevention of accidents in coal mines via safety organizations and unions. These results can provide guidelines for workers, industrialists, and government regarding occupational safety in the whole coal industry.

Owing to the rapid increase in construction and demolition (C&D) waste, the information of waste generation (WG) has been advantageously utilized as a strategy for C&D waste management. Recently, artificial intelligence (AI) has been strategically employed to obtain accurate WG information. Thus, this study aimed to manage demolition waste (DW) by combining three algorithms: artificial neural network (multilayer perceptron) (ANN-MLP), support vector regression (SVR), and random forest (RF) with an autoencoder (AE) to develop and test hybrid machine learning (ML) models. As a result of this study, AE technology significantly improved the performance of the ANN model. Especially, the performance of AE (25 features)–ANN model was superior to that of other non-hybrid and hybrid models. Compared to the non-hybrid ANN model, the performance of AE (25 features)–ANN model improved by 49%, 27%, 49%, and 22% in terms of the MAE, RMSE, R², and R, respectively. The hybrid model using ANN and AE proposed in this study showed useful results to improve the performance of the DWGR ML model. Therefore, this method is considered a novel and advantageous approach for developing a DWGR ML model. Furthermore, it can be used to develop AI models for improving performance in various fields.