A hybrid method for improved stability prediction in construction projects: a case study of stope hangingwall stability

New paper published on Applied Soft Computing

Paper Title: A hybrid method for improved stability prediction in construction projects: a case study of stope hangingwall stability

Authors: Chong-chong Qi*, Andy Fourie, Guo-wei Ma and Xiao-lin Tang

Abstract: 

Artificial intelligence (AI) approaches have proliferated in stability prediction of construction projects in the paste decade. However, the application of AI approaches did not reach the peak of its potential due to the inappropriate handling of missing data and the omission of state-of-the-art techniques. In the present contribution, we proposed a hybrid method for the improved stability prediction of construction projects based on individual machine learning (ML) algorithms, input missing data imputation, semi-supervised learning and the classifier ensemble. Seven ML algorithms were selected to build individual classifiers for the classifier ensemble. 5-fold cross validation was used as the validation method and the performance measures were chosen to be the confusion matrix, the receiver operating characteristic (ROC) curve and the area under ROC curve (AUC). Exhaustive grid search and firefly algorithm were used for hyper-parameters and weights tuning respectively. The capability of the proposed method was verified using an underground construction dataset, the stope hangingwall (HW) dataset. The case study shows that the input missing data imputation and semi-supervised learning improved the predictive performance of ML algorithms on HW stability prediction. The highest and average AUC values on the testing set were increased to 0.954 and 0.923 respectively on the expanded dataset, compared with 0.879 and 0.860 on the original complete dataset. Further improvement was obtained through the classifier ensemble, with the AUC value being increased to 0.976. Harnessing such method extends recent efforts for stability prediction in construction projects, and can significantly accelerate the project design and stability management.

Keywords: Hybird method; stability prediction; missing data imputation; semi-supervised learning; classifier ensemble.

Comments from reviews:

Reviewer 1: 

The paper presents novel hybrid method for improved stope hanging wall (HW) stability prediction, using input missing data imputation, semi-supervised learning and classifier ensemble. The method has been validated using an underground construction dataset, the stope HW stability dataset.

Seven well known machine learning algorithms were used and  classifier ensemble was  built on their optimal individual classifiers. The weighted voting approach was used as the ensemble method. The performance comparison was carried out  using the confusion matrix, the ROC curve ( (receiver operating characteristic curve) and the AUC (area under ROC curve). As a method for validation was used 5-fold cross-validation. Exhaustive grid search and the "firefly -algorithm" were used for hyper parameters and  weights tuning respectively. This method can significantly accelerate the project design and stability management.

My recommendation is this paper to be accepted for publication with minor corrections.

Reviewer 2: 

A hybrid method for improved risk estimation of construction projects based on input missing data imputation is suggested. Some ML algorithms and classifier ensemble technique are constructed. From a methodological view, there is no big contribution to soft computing. The paper needs a big revision.

Representative figure from this paper:

Flow-chart of the proposed hybrid method.

Read paper online.

Full paper can be obtained through Email.