Abstracts – Browse Results

Search or browse again.

Click on the titles below to expand the information about each abstract.
Viewing 4 results ...

Hayden, I (2019) An evaluation of the design and use of applied visual interactive resources for teaching building regulations in higher education built environment programmes. Architectural Engineering and Design Management, 15(03), 159–80.

Javed, S, Ørnes, I R, Myrup, M and Dokka, T H (2019) Design optimization of the borehole system for a plus-Energy kindergarten in Oslo, Norway. Architectural Engineering and Design Management, 15(03), 181–95.

Osei-Kyei, R and Chan, A P C (2019) Model for predicting the success of public–private partnership infrastructure projects in developing countries: a case of Ghana. Architectural Engineering and Design Management, 15(03), 213–32.

Tang, B, Han, J, Guo, G, Chen, Y and Zhang, S (2019) Building material prices forecasting based on least square support vector machine and improved particle swarm optimization. Architectural Engineering and Design Management, 15(03), 196–212.

  • Type: Journal Article
  • Keywords: Improved particle swarm optimization (IPSO); least square support vector machine (LSSVM); building materials; price prediction; parameter optimization;
  • ISBN/ISSN: 1745-2007
  • URL: https://doi.org/10.1080/17452007.2018.1556577
  • Abstract:
    In view of the uncertainty problems caused by the price fluctuation of building materials, this paper proposes a material price forecasting method based on least square support vector machine (LSSVM) and improved particle swarm optimization (IPSO). After training and testing the price prediction model with LSSVM, it is adopted to forecast the future material price. It can obtain the regularization parameter and kernel function parameter of LSSVM with the PSO global search ability, realizing the optimized selection of the LSSVM model parameters, which makes such prediction more close to actual value. Considering that the optimization process of standard PSO is easily trapped into local optimum and premature convergence, the average particle distance and fitness variance are introduced to improve the PSO. Finally, simulation analysis is carried out by using MATLAB, in addition, the online and offline performances of the algorithm are evaluated. Actual calculation examples show that the average relative error of the IPSO-LSSVM forecasting is 2.11%, and the root mean square relative error is 2.44%. The proposed method has the advantages of high prediction precision, fast convergence rate and good generalization ability compared with other traditional algorithms such as neural network and time series prediction. Therefore, it can accurately forecast the building material prices, reasonably determine the best purchase time, and has actual application value.