PSI - Issue 70

Abdul Musavvir et al. / Procedia Structural Integrity 70 (2025) 432–439

434

2.1. Decision Tree Regression: Decision Tree Regression which is a non-linear ML technique is used for predicting continuous values. It works by splitting data into subsets based on decision rules, minimizing prediction error. Its intuitive tree structure enhances interpretability, making it popular for tasks like price prediction and other regression applications. 2.2. Linear Regression: Linear regression is a model used to the analyze relationship between subordinate variable and autonomous factors. It predicts continuous outcomes by fitting a linear equation to observed data, allowing for insights into relationships and facilitating predictions based on input features. 2.3. Random Forest Regression: Linear regression is a model used to the analyze relationship between subordinate variable and autonomous factors. It predicts continuous outcomes by fitting a linear equation to observed data, allowing for insights into relationships and facilitating predictions based on input features. 2.4. XGBoost Regression: XGBoost, or Extreme Gradient Boosting, is a powerful open-source ML library designed for efficient and scalable gradient boosting. It excels in classification, regression, and ranking tasks by combining multiple weak models into a strong predictive model. XGBoost is widely used in data science competitions due to its high performance and speed. The predictive model for this research uses all four algorithms (Linear Regression, Decision Tree Regression, Random Forest and XGBoost Regression). The comparative analysis between these models has been done using the validated with the real time experimentation values. 3. Data Collection Data is crucial in ML as it serves as the foundation for training models. High-quality, relevant data enables algorithms to learn patterns and make accurate predictions. Insufficient or poor-quality data can lead to unreliable models, resulting in inaccurate outcomes. Thus, the effectiveness of ML systems heavily relies on the quantity and quality of the data used during training. A dataset of 355 (Abobakr Khalil Al-Shamiri, 2019) (Elshekh, Shafiq, Nuruddin, & Fathi., 2013) (Huu-Bang Tran, 2024) (Ibrahim Y. Hakeem, 2023) (Yeh., 1998) (Naraindas Bheel, 2024) (Sajjad Shokouh, 2025) (V.V. Praveen Kumar, 2023) (Yansheng Liu, 2024) (Yingqing Lyu, 2024) data for the 28 th day compressive strength of HSC with different mix combinations that including the following constituents Concrete, Super plasticizer, Fly Ash, Ground Granulated Blast Furnace Slag, Silica Fumes and Nano Silica has been extracted. All the data were extracted from reputed journals to maintain data quality.

Table 1. The description of extracted data

Cement (Kg/m 3 )

FA (Kg/m 3 )

CA (Kg/m 3 )

Water (Kg/m 3 )

SP (Kg/m 3 )

FA (Kg/m 3 )

GGBS (Kg/m 3 )

SF (Kg/m 3 )

NS (Kg/m 3 )

Compressive Strength (MPa)

355

355

355

355

355

355

355

355

355

Count

355

Mean 406.770 759.234 952.698 165.411 4.855

37.604 67.741

38.932 63.717

2.414

1.830

63.011 10.792 45.300 55.900 60.800

0.00 0.00 0.00 0.00

0.00 0.00 0.00 0.00

Std

110.975 95.937 153.063 19.993 125.000 552.000 644.000 125.000

6.282

0.00 0.00 0.00

0.00 0.00 0.00

Min

0.00 0.00 0.00

25% 325.800 698.000 852.100 155.650 50% 425.000 750.000 914.000 167.000