PREDICTION OF NANOFLUID SPECIFIC HEAT CAPACITY USING SUPERVISED REGRESSION MODELS

Simple item page
dc.contributor.authorYomna Abu Amra
dc.contributor.authorAshraf Y. Maghari
dc.date.accessioned2026-09-14T06:27:31Z
dc.date.available2026-09-14T06:27:31Z
dc.date.issued2026-05-12
dc.description.abstractBackground: Accurate prediction of the specific heat capacity (SHC) of nanofluids is essential for improving the design and performance of thermal management, heat transfer, and energy systems. Traditional experimental methods for measuring SHC are often costly, time-consuming, and prone to uncertainties caused by factors such as particle agglomeration, sedimentation, and temperature sensitivity. Therefore, reliable and efficient predictive approaches are needed. Recent advances in machine learning have shown strong potential in modeling complex thermophysical properties of nanofluids and providing fast, cost-effective alternatives to experimental measurements. Objectives: This study aimed to develop and evaluate supervised machine learning regression models for predicting the specific heat capacity (SHC) of nanofluids. The models were designed to use key thermophysical input parameters, including nanoparticle type, base fluid, nanoparticle volume fraction, and base fluid temperature, to provide an accurate, reliable, and practical predictive tool for engineering and energy applications. Methods: A dataset of 517 nanofluid records was collected from the Kaggle platform and preprocessed for analysis. Each record included four input features: nanoparticle type, base fluid, nanoparticle volume fraction, and base fluid temperature, while SHC (J/kg·K) was used as the target variable. Categorical variables were encoded using one-hot encoding, and correlation analysis was conducted to identify weak or redundant features. The dataset was split into 70% training and 30% testing sets. Four supervised regression models were developed and compared: Gradient Boosting, XGBoost, AdaBoost, and Decision Tree Regressor. Model performance was evaluated using Mean Squared Error (MSE), Mean Absolute Error (MAE), and the coefficient of determination (R²). Additionally, 5-fold cross-validation was applied to assess model robustness and reduce overfitting.
dc.identifier.urihttps://dspace.alquds.edu/handle/20.500.12213/10763
dc.language.isoen
dc.publisherDeanship of Scientific Research - Al-Quds University
dc.titlePREDICTION OF NANOFLUID SPECIFIC HEAT CAPACITY USING SUPERVISED REGRESSION MODELS
dc.typeArticle
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
PREDICTION OF NANOFLUID SPECIFIC HEAT CAPACITY USING SUPERVISED.pdf
Size:
97.57 KB
Format:
Adobe Portable Document Format
Description:
Download
License bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
license.txt
Size:
1.61 KB
Format:
Item-specific license agreed upon to submission
Description:
Download
Collections
Electrical, Architectural Engineering and Computer Science, Information Technology and Physics