www.nature.

com/scientificreports

OPEN Predicting compressive strength

of RCFST columns under different
loading scenarios using machine
learning optimization
Feng Wu 1, Fei Tang 1*, Ruichen Lu 2 & Ming Cheng 2
Accurate bearing capacity assessment under load conditions is essential for the design of concrete-
filled steel tube (CFST) columns. This paper presents an optimization-based machine learning method
to estimate the ultimate compressive strength of rectangular concrete-filled steel tube (RCFST)
columns. A hybrid model, GS-SVR, was developed based on support vector machine regression (SVR)
optimized by the grid search (GS) algorithm. The model was built based on a sample of 1003 axially
loaded and 401 eccentrically loaded test data sets. The predictive performance of the proposed model
is compared with two commonly used machine learning models and two design codes. The results
obtained for the axial loading dataset with ­R2 of 0.983, MAE of 177.062, RMSE of 240.963, and MAPE
of 12.209%, and for the eccentric loading dataset with ­R2 of 0.984, MAE of 93.234, RMSE of 124.924,
and MAPE of 10.032% show that GS-SVR is the best model for predicting the compressive strength
of RCFST columns under axial and eccentric loadings. It is an effective alternative method that can
be used to assist and guide the design of RCFST columns to save time and cost of some laboratory
experiments. Additionally, the impact of input parameters on the output was investigated.

An infill component known as a "concrete-filled steel tube" (CFST) is a structural system consisting of an outer
steel tube and a core filled with ­concrete1. The most commonly used types of CFST columns are circular con-
crete-filled steel tube (CCFST) and rectangular concrete-filled steel tube (RCFST), which effectively utilize the
complementary action between concrete and steel. Compared to conventional reinforced concrete or pure steel
elements, the CFST system provides mechanical advantages due to the steel tube’s restraining effect on the filled
concrete, substantially improving ductility and s­ trength2–7. Additionally, the concrete core restrains the inward
deformation of the steel tube, retarding local buckling and enhancing overall column s­ tability8,9. These synergistic
effects lead to increased strength and performance characteristics over the respective individual parts. Due to
their high strength, resilience, effective seismic energy absorption, and excellent fire resistance, CFST columns
are commonly used in high-rise buildings, bridges, and other infrastructure p ­ rojects10,11 .
The primary mechanical characteristic of CFST is its compressive strength, which plays a critical role in the
accurate design of CFST columns to ensure structural stability. To better understand the behavior of CFST under
loading, researchers commonly use experimental and finite element methods to estimate their ­performance12,13.
Physical experiments can provide valuable insights, but they are resource-intensive and time-consuming. On
the other hand, finite element analysis can reduce the number of tests required through computer simulation,
but its accuracy depends heavily on the expertise of the modeler and requires high computer configurations.
To address these limitations, some countries have developed equation-based design standards, such as ACI 318
(ACI 2014), Eurocode 4 (CEN 2004), AISC 360-16 (AISC 2016), and Chinese codes (GB 50936-2014 and GB/T
51446-2021), which are based on extensive experimental results. Using design codes to predict compressive
strength is currently a more practical option than physical experiments or finite element analysis. However, it is
important to note that these empirical formulas have their corresponding scope of application and may not be
suitable for all CFST columns, which vary in material strength, shape, cross-sectional length, and slenderness.
Therefore, using these standards to calculate the strength of CFST columns may carry a certain level of risk, and
additional caution and analysis may be required.
Machine learning techniques have the potential to provide accurate and efficient predictions of the bearing
capacity of CFST c­ omponents14–18. These methods can utilize large volumes of experimental data to identify

1
School of Architectural Engineering, Xinyang Vocational and Technical College, Xinyang 464000, China. 2China
Construction Fifth Engineering Division Corp., Ltd., Changsha 410000, China. *email: [email protected]

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 1

Vol.:(0123456789)
 www.nature.com/scientificreports/

patterns and relationships that are difficult to detect using traditional methods. Some of the machine learning
methods that have been used for this purpose include artificial neural networks (ANN), gene expression pro-
gramming (GEP), back-propagation neural networks (BPNN), and fuzzy logic. By leveraging these techniques,
researchers have been able to successfully predict the carrying capacity of CFST, which can provide valuable
insights for designing structures and reducing the need for further testing. Overall, the application of machine
learning to CFST design represents an exciting and promising area of r­ esearch19–28. In order to implement the
ultimate compressive strength prediction of RCFST columns, Mai et al.29 developed an ANN network that was
optimized by the particle swarm optimization algorithm. The results revealed that the proposed hybrid model
has higher prediction accuracy than the traditional design codes. The BAS-MLP model was created by Ren et al.30
using a multilayer perceptron (MLP) neural network coupled with a beetle antenna search (BAS) algorithm to
forecast the ultimate bearing capacity of RCFST columns. The outcomes demonstrated that the BAS-MLP model
performs better than a number of benchmark models and traditional approaches. To forecast the maximum
load capacity of short rectangular columns of restrained reinforced concrete (SCFST), Lu et al.31 established a
predictive method based on the gradient boost regression tree (GBRT) model. The results of a straightforward
comparison of many regression models revealed that the GBRT model makes a fair prediction of the mechanical
characteristics of SCFST columns. The ANN-PSO model was used by Kim et al.29,32 to forecast the eccentric load
capacity of 241 CCFST columns and 622 RCFST columns, respectively. The findings revealed that the average
prediction errors were 12.1% and 15.4%, respectively, which is better than the traditional design codes. On the
basis of 1224 test data, Panagiotis et al.33 developed an ANN model for the ultimate compressive capacity of
RCFST columns with seven variables, including the column’s width and height, steel tube thickness, effective
length, steel yield strength, concrete compressive strength, and eccentricity. They then compared the developed
model with the design codes currently in use. It was revealed that its accuracy was greatly enhanced while keeping
the forecast findings steady. Also, an explicit equation is provided for simple implementation and use evaluation.
Quang et al.34 developed a gradient tree boosting approach to forecast the strength of the CFST column, and the
proposed model produced higher prediction accuracy when compared to deep learning, decision trees, random
forests, and support vector machines (SVM).
Research on predicting the strength of CFST columns using machine learning seems to have made some
progress. However, most studies have focused on using traditional machine learning models to forecast the
axial compression strength of CFST columns. These models are limited by the selection of hyper-parameters,
resulting in restricted prediction accuracy. Optimized hybrid models have the potential to improve prediction
performance, but there is limited research in this area and further studies are necessary. Furthermore, current
research primarily focuses on load-bearing capacity predictions, with less emphasis on the feature importance
analysis of design parameters, which is particularly valuable for CFST design. To achieve this objective, this
study aims to establish an optimization model for the compressive strength of RCFST under axial and eccentric
loading conditions and analyze the impact of these design parameters on the output results.
As shown in Fig. 1, the input parameters consist of both geometric features and material properties. For
RCFST, the specific input variables include column width (B), height (H), thickness (T), length (L), yield strength
(fy), compressive strength (fc), top eccentricity (et), and bottom eccentricity (eb). The performance of the pro-
posed optimization model was compared with that of conventional support vector regression (SVR) and random

Figure 1.  Schematic diagram of RCFST columns under axial and eccentric loading.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 2

Vol:.(1234567890)
www.nature.com/scientificreports/

forest (RF) models to determine the optimal prediction model for this research. Moreover, the Shapley addi-
tive explanations (Shap) analysis method has been introduced to assess the roles and impacts of these design
parameters.

Methodology
Random forest model
The random forest algorithm is a machine learning method that combines decision trees, random feature selec-
tion, and integration to create a powerful combinatorial classifier. It uses a self-help approach to perform boot-
strap sampling and generate training subsets, ensuring that n random samples produce n training sets of the
same size.
Each training subset constructs its own decision tree separately, with the decision tree construction compris-
ing two processes: node splitting and random selection of feature variables. Node splitting is based on splitting
rules that compare information attributes and select the attributes with the best comparison results to generate
subtrees for growing the decision tree. As depicted in Fig. 2, random feature variable generation is commonly
used for the random selection of input variables and information attributes for node splitting. Random selec-
tion of training subsets and node attributes ensures the randomness of the random forest to prevent the model
from falling into the dilemma of overfitting and local over-optimization. Finally, the average of n decision tree
regression prediction results is chosen as the final prediction value.

Support vector regression model

Support vector regression (SVR) is based on the idea of structural risk minimization and is known for its good
performance and predictability when dealing with situations involving small samples, nonlinearities, and large
dimensions, as illustrated in Fig. 3. The basic concept behind SVR is to use nonlinearity to map the original
data x to a high-dimensional feature space, where the linear regression problem can be solved. The regression
function of SVR is shown below.
f (x) = w · φ(x) + b (1)
where w is the weight vector, b is the bias, and the following functions can be used to determine w and b.
 1 �n

 Minimize w T + c (ξi + ξi∗ )


 2 i=1
yi − w T φ(x) − b ≤ ε + ξi

(2)
 w T φ(x) + b − yi ≤ ε + ξi∗




ξi , ξi∗ ≥ 0


Figure 2.  Schematic diagram of random forest ­algorithm35.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 3

Vol.:(0123456789)
 www.nature.com/scientificreports/

Figure 3.  The schematic diagram of SVR.

where c is the penalty parameter, ξi and ξi∗ are slack variables, and ε is the insensitive range. There are various
options for the kernel function, and the typical RBF kernel function is utilized in this study. The calculation
process of SVR can be represented by the flow chart in Fig. 4.
For regression modeling, the interplay between two hyper-parameters (c and g) has the greatest impact on
model ­accuracy36. To address this issue, the grid search (GS) method is introduced, which is widely adopted due
to its ease of use and ­simplicity37.

Figure 4.  Algorithmic procedure of the ­SVR35.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 4

Vol:.(1234567890)
www.nature.com/scientificreports/

Support vector regression with grid search optimization

GS’s fundamental tenet is to first define the parameter area to be searched, split the region into a grid, and then
examine all possible parameter combinations at each intersection point in the grid. All of the grid’s intersections
represent parameter combinations (c, g) that must be searched, and all of the hyperparameter combinations must
be taken. Cross-validation is used to verify the prediction accuracy related to each set of data in order to get the
best (c,g). The sets of (c,g) with the best accuracy are chosen as the model’s core components. The basic steps of
grid parameter optimization search are as follows.

(1) Establish the coordinate grid: take x = [−a, a],y = [−b, b], step size L, and take the grid points of parameters
as c = ­2x, g = ­2y.
(2) Use k-fold cross-validation to find the regression accuracy: select the training data and divide them into k
copies that are uniformly disjoint, select k − 1 of them for model building, and leave the remaining one for
validating the model. A set (c,g) in the parameter grid is selected and the prediction accuracy of the test
data corresponding to this set (c,g) is recorded. Repeat the preceding processes k times to get k models,
then run each model on a different set of test data to get k prediction accuracies. Finally, take the average
of these accuracies to get the final corresponding accuracy of the group of parameters.
(3) Iterate the coordinate grid: find the final accuracy of all parameter combinations and rank them from largest
to smallest, and select the top group as the final (c,g) combination of the model.

Based on the results of parameter optimization and cross-validation, the best combination of hyper-param-
eters values is selected to make the system perform best, and the test dataset prediction is implemented using
SVR model with optimal parameters. The framework of this paper is shown in Fig. 5.

Tests of RCFST Tests of RCFST

under axial under eccentric
loading loading

Dataset

Set the parameter

range and step size
Training set

Create a grid with

Parameters c parameters c and g
and g
Test set
Search each grid node
SVR model with
cross-validation
Update parameters c
and g N
Meet the criteria?
Grid search process
SVR model with best
Y
parameters combination

Evaluation indexes: Evaluation of model SVR model and random

R2, MAE, MAPE,RMSE performance forest model

Feature importance
analysis

Figure 5.  The framework of this paper.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 5

Vol.:(0123456789)
 www.nature.com/scientificreports/

Dataset description
To construct a precise strength model for the CCFST column, a comprehensive experimental database is essential.
Two datasets, comprising 1003 tests on RCFST columns under axial loading (Dataset 1) and 401 tests on RCFST
columns under eccentric loading (Dataset 2), were collected from an open-public d ­ ataset38. A description of
the experimental conditions and more detailed experimental situations for each sample in the data set can be
found in R ­ eference39 and will not be repeated here. The ranges and statistical characteristics of these datasets are
illustrated in Fig. 6 and Table 1, respectively. It is noteworthy that the distribution of maximum bearing capacity
exhibits significant variations, which may pose a challenge to accurately predict the outcomes.
Also, it can be observed from Fig. 7 that Pearson linear correlations were computed and plotted between the
input and output variables in the two data sets. It can be seen from Fig. 7 that the three variables with the strongest
linear correlation with the compressive strength of RCFST are B, H, and T, which are all geometric properties.
The correlation coefficients of these three variables were 0.65, 0.56, and 0.56 for dataset 1 and 0.55, 0.60, and 0.50
for dataset 2, respectively. And the three of them are positively correlated with the compressive strength, while
the L is negatively correlated with the compressive strength. Among the multiple variables listed in this paper,
all the parameters except L, et, and eb are positive for the bearing capacity of RCFST columns, and N increases
as these parameters increase. However, the correlation coefficient between input and output variables did not
exceed 0.8, indicating that complex nonlinear correlations need to be established between multiple input factors
and the output compressive strength to achieve an accurate prediction of compressive strength.
Additionally, the following four metrics were used to evaluate the model’s performance: correlation coefficient
­(R2), root mean square error (RMSE), mean absolute error (MAE), and mean absolute percentage error (MAPE).
Their definitions are depicted b ­ elow40,41.
n
(T − T)(Y − Y ))2

(
2 i=1
R = n n (3)
2
(Y − Y )2

(T − T)
i=1 i=1


 n
1 
RMSE =  (T − Y )2 (4)
n
i=1

n
1
MAE= (T − Y ) (5)
n
i=1

n  
100   T − Y 
MAPE= (6)
n  Y 
i=1

where T and Y are the experimental and predicted results, respectively, while T and Y are the mean values.

Results and analysis

Optimization of optimal hyper‑parameter combination
The training set and the data set were chosen randomly for each case, with a ratio of 80%:20% between the two.
The cross-validation and GS method were used to explore the optimal hyper-parameter combination. The evo-
lution of the mean square error (MSE) and the determination of the optimal parameters during the training in
the search ­range2–5,25 are shown in Fig. 8. For dataset 1, the best validation performance of the model is achieved
when c = 2, g = 0.87055, and for dataset 2, when c = 6.9644, g = 0.5. Then, these two sets of hyper-parameter com-
binations will be used for the model building of the two data sets respectively.

Model prediction outcomes comparison

Random forest and the original SVR model were also utilized on the same training and test sets for comparison to
evaluate the validity and reliability of the proposed models. Figure 9 shows the relationship between the experi-
mental data under various scenarios and the forecasted results of the three models. As seen, for both the training
and test sets, the scatter between the three machine learning models’ outcomes and actual values is primarily
within ± 20%. Unfortunately, Fig. 9 makes it challenging to compare the three models. The error metrics between
the predicted outcomes and the actual values of the various models are listed in Table 2 for easy comparison.
The correlations between the predicted and actual values in the hybrid model proposed in this research are
0.983 and 0.984 for two different datasets, respectively, which are higher than those in the two standard machine
learning models, RF and SVR. Among the three models, the other three error indicators are the lowest. The
­ 2 of 0.983, MAE of 177.062, RMSE of 240.963, and MAPE of
results obtained for the axial loading dataset with R
12.209%, and for the eccentric loading dataset with R ­ 2 of 0.984, MAE of 93.234, RMSE of 124.924, and MAPE of
10.032% show that GS-SVR is the best model for predicting the compressive strength of RCFST columns under
axial and eccentric loadings.
Figure 10 offers a comprehensive overview of the prediction error distribution among the models in the test
dataset. The findings reveal that, across all three machine learning models, approximately 50% of the test sets
exhibit a relative prediction error of 10% or less, while 80% of the test sets display a relative error distribution

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 6

Vol:.(1234567890)
www.nature.com/scientificreports/

13500 13800

9000 9200

N(kN)

N(kN)
4500 4600

0 0

100 200 300 400 0 100 200 300 400

B(mm) H(mm)

13800 13800

9200 9200

N(kN)
N(kN)

4600 4600

0 0

0 5 10 15 20 0 1000 2000 3000 4000 5000

T(mm) L(mm)

13800 13800

9200 9200
N(kN)

N(kN)

4600 4600

0 0

210 420 630 840 0 48 96 144

fy(MPa) fc(MPa)

(a) RCFST under axial loading

13500 13800

9000 9200
N(kN)

N(kN)

4500 4600

0 0

100 200 300 0 100 200 300 400

B(mm) H(mm)

Figure 6.  The range distribution of all variables.

within the 20% range. Moving to Fig. 11, it presents the prediction error statistics for the test set across various
operating conditions for each model. The optimized hybrid model demonstrates an average relative prediction

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 7

Vol.:(0123456789)
 www.nature.com/scientificreports/

13800 13800

9200 9200

N(kN)

N(kN)
4600 4600

0 0

0 5 10 15 20 0 1000 2000 3000 4000 5000

T(mm) L(mm)

13800 13800

9200 9200
N(kN)

N(kN)
4600 4600

0 0

200 400 600 800 0 56 112 168

fy(MPa) fc(MPa)

13500 13800

9000 9200
N(kN)

N(kN)

4500 4600

0 0

0 100 200 300 0 100 200 300 400

et(mm) eb(mm)
(b) RCFST under eccentric loading

Figure 6.  (continued)

Data set Size Variable Unit Min Max Median Mean SD Kurtosis Skewness
B mm 60 400 125 141.59 55.34 2.12 1.39
H mm 44 400 149.92 153.42 54.71 2.09 1.16
T mm 0.7 18.5 4.38 4.57 2.23 3.76 1.48
1 1003 L mm 60 4500 570 949.63 864.16 1.83 1.64
fy MPa 115 835 342 405.82 171.62 0.32 1.17
fc MPa 8.52 164.1 44.9 55.07 33.23 0.83 1.17
N kN 105.4 14,116 1697 2238.31 1919.5 4.99 2.03
B mm 76.2 323 150 150.52 46.19 2.5 1.28
H mm 76.2 323 150 154.56 49.57 1.58 1.12
T mm 1.9 12.5 4.18 4.51 1.65 5.97 1.96
L mm 330 4910 1800 1776.09 1118.3 −0.59 0.5
2 401 fy MPa 242 761 340 390.74 126.2 1.54 1.48
fc MPa 18.76 183 46.8 57.18 30.64 1.9 1.18
et mm 0.9 300 30 43.87 45.32 12.22 3.18
eb mm −25 300 30 40.79 47.01 11 2.95
N kN 156.35 7136 981.6 1250.12 1005.45 8.5 2.36

Table 1.  The data set’s statistical findings.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 8

Vol:.(1234567890)
www.nature.com/scientificreports/

B
1.0 B
B 1.0 H 0.80
1.0
B 1.0 H
0.80
H 0.88 1.0 T 0.60 H 0.71 1.0 T 0.60
0.40 T 0.15 0.18 1.0 L
T 0.13 0.095 1.0 L 0.40
0.20 L -0.26-0.30 0.21 1.0 fy 0.20
L 0.0074 0.023 -0.11 1.0 fy 0.0
fy 0.15 0.18 0.44-0.095 1.0 fc 0.0
-0.20
fy 0.078 0.034 0.43 0.062 1.0 fc fc3.1E-4-0.13 0.23 0.19-0.015 1.0 et -0.20
-0.40 -0.40
et 0.45 0.40 0.14 -0.24 0.15 -0.14 1.0 eb
fc -0.033 -0.096 0.40 -0.033 0.45 1.0 N -0.60 -0.60
eb 0.46 0.46 0.15 -0.29 0.17 -0.18 0.98 1.0 N
-0.80 -0.80
N 0.65 0.59 0.59 -0.12 0.53 0.49 1.0 N 0.55 0.60 0.50 -0.18 0.46 0.35-0.065-0.016 1.0
-1.0 -1.0

(a) Dataset 1 (b) Dataset 2

Figure 7.  Pearson correlation coefficient of variables.

Figure 8.  Optimal hyper-parameter combination search using grid search and cross-validation.

error of 12.209% and 10.032% for the test set under the two different working conditions, respectively. These
average relative errors are notably smaller than those of the corresponding SVR and random forest models,
with all relative errors falling within the 15% threshold, meeting the requirements for engineering applications.
To further evaluate the performance of the proposed model, two design criteria, AISC 360–16 and Eurocode
4 (EC4), were used to make predictions on the test set and the ratio between the experimental and predicted
values of the different models was calculated as shown in Fig. 12. From the mean values μ presented in Fig. 12,
the ratio between the actual and predicted values in the GS-SVR model is closer to 1, indicating that the predic-
tions are more accurate.

Input feature analysis

In addition to accurate load-bearing capacity predictions, the analysis of the importance of design parameters
is also a critical step in the design of RCFST columns. This is because adjusting design parameters in order of
importance, from high to low, can save time and costs. This section introduces Shap analysis to discuss the impact
of various parameters on the output results, as shown in Fig. 13. The factors that have the greatest impact on the
load-bearing capacity of the column are, in descending order of importance, H, followed by B, and then fy, T, fc,
and L. The eccentricities et and eb have the least impact. Additionally, Fig. 13 also demonstrates whether these
impacts are positive or negative. It can be observed that the top five parameters in terms of importance have a
positive impact on compressive strength, while length and eccentricity have a negative impact. These influences
are extremely helpful in the design of RCFST columns. Designers can adjust the design values of various param-
eters based on the impact of these design parameters to achieve the desired design objectives.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 9

Vol.:(0123456789)
 www.nature.com/scientificreports/

15000 15000
SVR y=1.2x
SVR y=1.2x
RF RF
GS-SVR GS-SVR

Predicted value
10000 10000

Predicted value
y=0.8x y=0.8x

5000 5000

0 0
0 5000 10000 15000 0 5000 10000 15000
Experimental value Experimental value
(a) RCFST under axial loading:training set (left) and test set (right)

8000 8000
SVR y=1.2x
SVR y=1.2x
RF RF
GS-SVR GS-SVR
6000 6000

Predicted value
Predicted value

y=0.8x y=0.8x
4000 4000

2000 2000

0 0
0 2000 4000 6000 8000 0 2000 4000 6000 8000
Experimental value Experimental value
(b) RCFST under eccentric loading:training set (left) and test set (right)

Figure 9.  Correlation between expected results and actual values.

Dataset 1 Dataset 2
Model Evaluation indices Training Test Training Test
R2 0.981 0.980 0.985 0.978
MAE 188.123 194.715 94.504 116.753
SVR
RMSE 266.234 262.589 124.929 147.148
MAPE(%) 12.127 13.837 11.260 13.595
R2 0.985 0.980 0.986 0.979
MAE 164.995 195.491 89.305 108.293
RF
RMSE 239.469 259.248 119.205 141.457
MAPE(%) 10.295 13.400 10.794 13.440
R2 0.988 0.983 0.987 0.984
MAE 144.931 177.062 85.270 93.234
GS-SVR
RMSE 213.687 240.963 114.183 124.924
MAPE(%) 9.420 12.209 10.066 10.032

Table 2.  Evaluation indicators for the three models’ predictions.

Conclusions
This study proposes an optimal hybrid model to accurately predict the strength of RCFST columns under both
axial and eccentric loads, shedding light on the complex mechanical behavior of RCFST. The proposed model
considers the intricate interactions between geometry, material properties, and compressive strength for various
loading scenarios.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 10

Vol:.(1234567890)
www.nature.com/scientificreports/

100 4% >30%
7% 7% 6% 8% 11%
7% 20%-30%
11% 11% 16%
14% 6% 15%-20%
80 9%
11% 5% 10%-15%
11% 11%

Cumulative percentage
12% 5%-10%
19% 11%
15% <5%
60 17% 18%
18%
21%
23%
22% 24%
40 24%
24%

20 43%
31% 35% 33%
24% 29%

0
SVR RF GS-SVR SVR RF GS-SVR
Dataset 1 Dataset 2

Figure 10.  Prediction error distribution of the test set.

50
25%~75% 1.5IQR Median Average Outliers

40
Relative prediction error/%

30

20

10

SVR RF GS-SVR SVR RF GS-SVR

Dataset 1 Dataset 2

Figure 11.  Box plot of prediction error distribution of test set.

4 4
GS-SVR GS-SVR
EC4 EC4
AISC 360-16 AISC 360-16
3 3
Test/predicted ratio

Test/predicted ratio

GS-SVR: μ=1.018 GS-SVR: μ=1.017

EC4: μ=1.077 EC4: μ=1.097
AISC 360-16: μ=1.123 AISC 360-16: μ=1.152
2 2

1 1

0 0
0 50 100 150 200 0 20 40 60 80
Test sample Test sample

(a)Dataset 1 (b) Dataset 2

Figure 12.  Ratio of experimental values to predicted values for different models.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 11

Vol.:(0123456789)
 www.nature.com/scientificreports/

High
H
H
B
B
fy
fy

Feature value
T
T
fc
fc
L
L
et
et
eb
eb
Low
0 50 100 150 200 250 300 350 -500 0 500 1000 1500 2000
M ean(|SHAP value|)(average impact on model output magnitude) SHAP value(impact on model output)

Figure 13.  SHAP feature importance and summary plot for RCFST column under eccentric loading.

For two different test sets, the suggested hybrid model exhibits average relative prediction errors of 12.209%
and 10.032%, respectively. These errors are smaller than those of the traditional SVR and random forest models,
and all relative errors are under 15%, indicating a high degree of prediction accuracy. Moreover, the proposed
hybrid model has certain superiorities over the traditional design codes. Therefore, the optimal hybrid model can
serve as a reliable alternative to commonly used design codes for predicting the compressive strength of RCFST
columns, which can partially replace laboratory tests to save resources and assist in the design of RCFST columns.
Among the input parameters listed in this paper, the cross-sectional dimensions of the steel tube concrete
are the most influential on its compressive strength. In the design of concrete-filled steel tube columns, attention
should be given to the width and height of the RCFST column. Parameters et, eb, and L have a negative effect on
compressive strength, while other geometric parameters and material properties lead to an increase in compres-
sive strength with an increase in their design values.
The implementation of the proposed model in this paper is on a specific dataset. The applicability and gener-
alizability to other similar datasets need to be further investigated. Also, taking more factors affecting the bearing
capacity into account as variables within the model is a focus for future work.

Data availability
The datasets used and/or analyzed during the current study available from the corresponding author on reason-
able request.

Received: 1 April 2023; Accepted: 24 September 2023

References
1. Güneyisi, E. M., Gültekin, A. & Mermerdaş, K. Ultimate capacity prediction of axially loaded CFST short columns. Int. J. Steel
Struct. 16(1), 99–114 (2016).
2. Portolés, J. M., Serra, E. & Romero, M. L. Influence of ultra-high strength infill in slender concrete-filled steel tubular columns. J.
Constr. Steel Res. 86, 107–114 (2013).
3. Xiong, M.-X., Xiong, D.-X. & Liew, J. Y. R. Axial performance of short concrete filled steel tubes with high- and ultra-high- strength
materials. Eng. Struct. 136, 494–510 (2017).
4. Ye, Y., Han, L.-H., Sheehan, T. & Guo, Z.-X. Concrete-filled bimetallic tubes under axial compression: Experimental investigation.
Thin-Walled Struct. 108, 321–332 (2016).
5. Ekmekyapar, T. & Al-Eliwi, B. J. M. Experimental behaviour of circular concrete filled steel tube columns and design specifications.
Thin-Walled Struct. 105, 220–230 (2016).
6. Lu, Y., Li, N., Li, S. & Liang, H. Behavior of steel fiber reinforced concrete-filled steel tube columns under axial compression. Constr.
Build Mater. 95, 74–85 (2015).
7. Xiong, M.-X., Xiong, D.-X. & Liew, J. Y. R. Behaviour of steel tubular members infilled with ultra high strength concrete. J. Constr.
Steel Res. 138, 168–183 (2017).
8. Li, D., Huang, Z., Uy, B., Thai, H.-T. & Hou, C. Slenderness limits for fabricated S960 ultra-high-strength steel and composite
columns. J. Constr. Steel Res. 159, 109–121 (2019).
9. Patel, V. I. et al. Ultra-high strength circular short CFST columns: Axisymmetric analysis, behaviour and design. Eng. Struct. 179,
268–283 (2019).
10. Han, L.-H., Li, W. & Bjorhovde, R. Developments and advanced applications of concrete-filled steel tubular (CFST) structures:
Members. J. Constr. Steel Res. 100, 211–228 (2014).
11. Tao, Z., Han, L.-H. & Wang, D.-Y. Strength and ductility of stiffened thin-walled hollow steel structural stub columns filled with
concrete. Thin-Walled Struct. 46(10), 1113–1128 (2008).
12. Du, Y., Chen, Z., Richard Liew, J. Y. & Xiong, M.-X. Rectangular concrete-filled steel tubular beam-columns using high-strength
steel: Experiments and design. J. Constr. Steel Res. 131, 1–18 (2017).
13. Nguyen, T.-T., Thai, H.-T., Ngo, T., Uy, B. & Li, D. Behaviour and design of high strength CFST columns with slender sections. J.
Constr. Steel Res. 182, 106645 (2021).

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 12

Vol:.(1234567890)
www.nature.com/scientificreports/

14. Barkhordari, M. S. & Massone, L. M. Ensemble techniques and hybrid intelligence algorithms for shear strength prediction of
squat reinforced concrete walls. Adv. Comput. Des. 8(1), 37–59 (2023).
15. Pham, V.-T. & Kim, S.-E. A robust approach in prediction of RCFST columns using machine learning algorithm. Steel Compos.
Struct. 46(2), 153–173 (2023).
16. Barkhordari Mohammad, S. & Billah, A. H. M. M. Efficiency of data-driven hybrid algorithms for steel-column base connection
failure mode detection. Pract. Period. Struct. Des. Constr. 28(1), 04022061 (2023).
17. Le, T.-T., Phan, H. C., Duong, H. T. & Le, M. V. Optimal design of circular concrete-filled steel tubular columns based on a com-
bination of artificial neural network, balancing composite motion algorithm and a large experimental database. Expert Syst. Appl.
223, 119940 (2023).
18. Nguyen, T.-A., Trinh, S. H., Nguyen, M. H. & Ly, H.-B. Novel ensemble approach to predict the ultimate axial load of CFST columns
with different cross-sections. Structures 47, 1–14 (2023).
19. Khalaf, A.A., Naser, K. Z. & Kamil, F. N. A. Predicting the Ultimate Strength of Circular Concrete Filled Steel Tubular Columns by
Using Artificial Neural Networks (2018).
20. Nour, A. I. & Güneyisi, E. M. Prediction model on compressive strength of recycled aggregate concrete filled steel tube columns.
Compos. Part B Eng. 173, 106938 (2019).
21. Tran, V.-L., Thai, D.-K. & Nguyen, D.-D. Practical artificial neural network tool for predicting the axial compression capacity of
circular concrete-filled steel tube columns with ultra-high-strength concrete. Thin-Walled Struct. 151, 106720 (2020).
22. Naser, M. Z., Thai, S. & Thai, H.-T. Evaluating structural response of concrete-filled steel tubular columns through machine learn-
ing. J. Build. Eng. 34, 101888 (2021).
23. Luat, N.-V., Shin, J. & Lee, K. Hybrid BART-based models optimized by nature-inspired metaheuristics to predict ultimate axial
capacity of CCFST columns. Eng. Comput. 38(2), 1421–1450 (2022).
24. Liao, J. et al. Novel fuzzy-based optimization approaches for the prediction of ultimate axial load of circular concrete-filled steel
tubes. Buildings-Basel 11, 12 (2021).
25. Ahmadi, M., Naderpour, H. & Kheyroddin, A. ANN model for predicting the compressive strength of circular steel-confined
concrete. Int. J. Civ. Eng. 15(2), 213–221 (2017).
26. Moon, J., Kim, J. J., Lee, T.-H. & Lee, H.-E. Prediction of axial load capacity of stub circular concrete-filled steel tube using fuzzy
logic. J. Constr. Steel Res. 101, 184–191 (2014).
27. Sarir, P., Chen, J., Asteris, P. G., Armaghani, D. J. & Tahir, M. M. Developing GEP tree-based, neuro-swarm, and whale optimiza-
tion models for evaluation of bearing capacity of concrete-filled steel tube columns. Eng. Comput. 37(1), 1–19 (2021).
28. Sarir, P. et al. Optimum model for bearing capacity of concrete-steel columns with AI technology via incorporating the algorithms
of IWO and ABC. Eng. Comput. 37(2), 797–807 (2021).
29. Nguyen, M. S. T. & Kim, S.-E. A hybrid machine learning approach in prediction and uncertainty quantification of ultimate
compressive strength of RCFST columns. Constr. Build. Mater. 302, 124208 (2021).
30. Ren, Q., Shen, Y. & Li, M. Hybrid intelligence approach for performance estimation of rectangular CFST columns under different
loading conditions. Structures 39, 720–738 (2022).
31. Lu, D., Chen, Z., Ding, F., Chen, Z. & Sun, P. Prediction of mechanical properties of the stirrup-confined rectangular CFST stub
columns using FEM and machine learning. In Mathematics. Vol. 9 (2021).
32. Nguyen, M.-S. T., Trinh, M.-C. & Kim, S.-E. Uncertainty quantification of ultimate compressive strength of CCFST columns using
hybrid machine learning model. In Engineering with Computers (2021).
33. Asteris, P. G., Lemonis, M. E., Le, T.-T. & Tsavdaridis, K. D. Evaluation of the ultimate eccentric load of rectangular CFSTs using
advanced neural network modeling. Eng. Struct. 248, 113297 (2021).
34. Vu, Q.-V., Truong, V.-H. & Thai, H.-T. Machine learning-based prediction of CFST columns using gradient tree boosting algorithm.
Compos. Struct. 259, 113505 (2021).
35. Migallón, V., Penadés, H., Penadés, J. & Tenza-Abril, A. J. A machine learning approach to prediction of the compressive strength
of segregated lightweight aggregate concretes using ultrasonic pulse velocity. In Applied Sciences. Vol. 13 (2023).
36. Wu, Y. & Li, S. Damage degree evaluation of masonry using optimized SVM-based acoustic emission monitoring and rate process
theory. Measurement 190, 110729 (2022).
37. Wu, Y. & Zhou, Y. Hybrid machine learning model and Shapley additive explanations for compressive strength of sustainable
concrete. Constr. Build. Mater. 330, 127298 (2022).
38. Thai, H.-T.T. et al. Concrete-filled steel tubular (CFST) columns database with 3,208 tests. Mendeley Data https://d ​ oi.o
​ rg/1​ 0.1​ 7632/​
j3f5c​x9yjh.1 (2020).
39. Thai, H.-T. et al. Reliability considerations of modern design codes for CFST columns. J. Constr. Steel Res. 177, 106482 (2021).
40. Wu, Y. & Zhou, Y. Prediction and feature analysis of punching shear strength of two-way reinforced concrete slabs using optimized
machine learning algorithm and Shapley additive explanations. Mech. Adv. Mater. Struct. 2022, 1–11 (2022).
41. Wu, Y. & Zhou, Y. Splitting tensile strength prediction of sustainable high-performance concrete using machine learning techniques.
Environ. Sci. Pollut. Res. 29(59), 89198–89209 (2022).

Author contributions
F.W.: writing-original draft. F.T.: methodology. R.L.: software. M.C.: software, formal analysis.

Competing interests
The authors declare no competing interests.

Additional information
Correspondence and requests for materials should be addressed to F.T.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional affiliations.

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 13

Vol.:(0123456789)
 www.nature.com/scientificreports/

Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. The images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/.

© The Author(s) 2023

Scientific Reports | (2023) 13:16571 | https://doi.org/10.1038/s41598-023-43463-6 14

Vol:.(1234567890)