Development of a High-Performance Concrete Compressive-Strength Prediction Model Using an Ensemble Machine-Learning Method Based on Bagging and Stacking (original) (raw)
Journal of the Computational Structural Engineering Institute of Korea
Predicting the compressive strength of high-performance concrete (HPC) is challenging because of the use of additional cementitious materials; thus, the development of improved predictive models is essential. The purpose of this study was to develop an HPC compressive-strength prediction model using an ensemble machine-learning method of combined bagging and stacking techniques. The result is a new ensemble technique that integrates the existing ensemble methods of bagging and stacking to solve the problems of a single machine-learning model and improve the prediction performance of the model. The nonlinear regression, support vector machine, artificial neural network, and Gaussian process regression approaches were used as single machine-learning methods and bagging and stacking techniques as ensemble machine-learning methods. As a result, the model of the proposed method showed improved accuracy results compared with single machine-learning models, an individual bagging technique model, and a stacking technique model. This was confirmed through a comparison of four representative performance indicators, verifying the effectiveness of the method.
Sign up for access to the world's latest research.
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Related papers
Suggestion for Concrete Strength Grade Using Ultrasonic Velocity for Tunnel Lining Concrete
Journal of the Korean Recycled Construction Resources Institute, 2015
This study implicates a quality evaluation technique for lining concrete using the ultra-sonic technique and proposed a classification for quality evaluation. From the comparison of test results, an improved result comparable to the results by destructive test can be obtained by using averaged transmission velocity of velocities obtained at pure space and central space between two sensors of ultrasonic tester. It is also shown that strength evaluation by ultra-sonic method is not reliable and an evaluation according to the transmission velocities can be improved method for the strength evaluation of concrete.
Journal of the Korean Institute of Building Construction, 2019
As a cold-weather-concrete construction technique for enhancing the sustainability and improving efficiency of cold-weather construction, the cracking timing, the starting point of deterioration for concrete, due to the shrinkage of the blast furnace slag cement concrete including accelerator was evaluated. As a result, by using blast furnace slag and accelerator, the cracking was developed faster with higher cracking potential under the restrained conditions at constant age and free-shrinkage strain. It can be considered that the results of decreased stress relaxation by creep or increased restraint with increased free-shrinkage strain causes the increased cracking development speed. Hence, it should be considered the necessary of cracking due to the shrinkage when blast furnace slag or accelerator was used for cold-weather construction.
Evaluation of Compression Molding Simulation with Compression Properties of Carbon Fiber Prepreg
Composites research, 2018
In order to optimize the prepreg compression molding (PCM) process, the forming simulation is required to cope with any problems that may be raised during the process. For the improvement of simulation accuracy, the input data of material property should be measured accurately. However, most studies assume that the compressive properties of the prepreg are identical to the tensile properties without quantifying them separately. Therefore, in this study, the in-plane compressive properties of the prepreg are presented to improve the accuracy of the forming simulation. As a result, the compressive modulus of the fibers was measured to be about 10-2 times lower than the tensile modulus. Also we designed a square-cup mold with a tilting angle of 110° to simulate the prepreg formability during the high temperature compression mold process. Shear angles were measured at each corner, which were compared with the simulation results. It was observed that the simulation results using the accurate compressive properties of the prepreg showed a similar trend with the experimental results. It was confirmed that the measured data of the in-plane compression property improved the accuracy of the forming simulation results.
A Study on P2P Lending Deadline Prediction Model based on Machine Learning
Journal of KIISE, 2019
Recently, there has been an increase in P2P lending users, a product that supports investments through lending among individuals using online platforms. However, since P2P lending's investors have to take financial risks, the investors may fail to investment due to the close of investment while they considering whether to invest or not. This paper predicts how long an investment product will take from a certain point to the close in order to provide deadline information for P2P loan investment products. To predicts the investment deadline, we have transforms into Timeseries data and Step data based on investment information on actual P2P products. The regression, classification, and time series prediction model were generated using machine learning algorithm. The results of the performance evaluation showed that in the Timeseries data-based model, the Multi-layer Perceptron regression model and the classification model showed the highest performance at 0.725 and 0.703 respectively. The Step data-based model was also the highest with the Multi-layer Perceptron regression model and the classification model at 0.782 and 0.651 respectively.
A Study of the Catenary Action for Axially Restrained Concrete Beams in Fire
Journal of the Korea Concrete Institute, 2018
The main research objective of this paper is to understand the load carrying mechanisms of reinforced concrete structures when subjected to fire. In particular, this research will focus on catenary action of reinforced concrete beams, which is one of the main strategies of providing structural robustness, in the form of the tying method. Catenary action can only be activated if there is axial restraint to the beam. The research was carried out using the commercial finite element package TNO DIANA. The behavior of an axially restrained beam goes through three stages: initial compression due to restrained thermal expansion, followed by almost pure bending when the axial force in the beam is very small, then transits to catenary action when the concrete is cracked and the tensile force in the reinforcement steel resists the applied catenary tensile action, until fracture of the tensile reinforcement steel.
Many researchers have reported that adding steel fiber to concrete improved its tensile and flexural strength significantly, but relatively few studies have been made on the compressive behavior of steel fiber-reinforced concrete. It is still less in case of high strength steel fiber-reinforced cementitious composites(SFRC). The main objective of this research is to examine the effect of adding steel fiber on the compressive strength of high strength SFRC using fiber reinforcing index(RI, Vf(lf/df)). It was found from the study that compressive strength was noticeably increased in proportion to RI. In conclusion, the relationship between Reinforcing Index(RI) and compressive strength in case of high strength steel fiber-reinforced cementitious composites was suggested.
Procedural Modeling Algorithm for Traditional Stone Fence Creator
Journal of Digital Convergence, 2013
In this paper, we present a procedural modeling algorithms to create Korean traditional stone fence using the fractal subdivision. The main process of the algorithm is to get the next step mesh by subdividing each triangle in the previous step triangular mesh. This process is repeated recursively. Dividing each triangle into four sub-triangles after choosing a random point on each side of the triangle and moving each vertices in the normal direction with random perturbations make the bumpy appearance of stone fences. In each step we remove flat vertices which does not influence the shape of the stone. The discrete curvature determines the flatness of a vertex. New triangles whose vertices are the vertices around the removed vertex are added to make a triangular mesh.
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.