Hyperparameter Tuning for Artificial Neural Network Pre­ dicting Concrete Compressive Strength

Bachelor Thesis (2021)
Author(s)

S.H. Seuntjens (TU Delft - Civil Engineering & Geosciences)

Contributor(s)

B. Šavija – Mentor (TU Delft - Civil Engineering & Geosciences)

Zhi Wang – Mentor (Tianjin University)

Faculty
Civil Engineering & Geosciences
More Info
expand_more
Publication Year
2021
Language
English
Graduation Date
25-06-2021
Awarding Institution
Delft University of Technology
Programme
Civil Engineering
Faculty
Civil Engineering & Geosciences
Downloads counter
310
Collections
thesis
Reuse Rights

Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.

Abstract

Concrete compressive strength is the most frequently used and most important mechanical property of concrete. National and international building codes (such as the Eurocode) frequently use compressive strength for design with concrete.

In some cases, instead of testing concrete specimens under compressive loading in laboratories, predicting the compressive strength using machine learning could be a good alternative. The ingredients making up the concrete mix and the curing age can be used as predictors for the compressive strength.

Artificial Neural Networks (ANNs) are machine learning algorithms that have been used since the nineteen sixties and were inspired by the way neurons work in the brain. Previous research suggests that ANNs have great potential to predict concrete compressive strength.

In this research, an Artificial Neural Network was set up using the Keras framework and was trained with concrete composition data consisting of examples of concrete recipes and their respective concrete compressive strengths. Then, three hyperparameter optimization methods (for-loop, grid search, and Bayesian optimization) were implemented in several runs. The resulting hyperparameters were used to create ANNs. Afterwards, the three methods were compared with respect to several metrics (R-squared score, root mean square error, and running time) to see which one is the relatively best method to provide hyperparameters for the predefined ANN that learns from concrete composition data.

The best runs of the three hyperparameter optimization methods show similar goodness-of-fit (with negligible differences). Among these best runs, grid search has the shortest running time. Bayesian optimization provides the highest R-squared score and has a root mean square error of 4.45 [MPa] and a reasonable running time of 73 minutes. Therefore, Bayesian optimization is considered the preferable hyperparameter optimization algorithm for the concrete compressive strength data, as goodness-of-fit is in most cases considered the decisive metric for this application.

Further research using larger parameter grids and search spaces, or other tuning methods, could result in even better goodness-of-fit results.

Files

BEP_Seuntjens_20210624.pdf
(pdf | 1.04 Mb)
License info not available