{"id":1645,"date":"2025-03-26T08:00:36","date_gmt":"2025-03-26T08:00:36","guid":{"rendered":"https:\/\/jjeci.net\/?p=1645"},"modified":"2025-04-13T09:30:38","modified_gmt":"2025-04-13T09:30:38","slug":"global-modeling-and-predicting-of-thermal-conductivity-of-nanofluids-using-an-ensemble-of-different-tree-based-gradient-boosting-algorithms","status":"publish","type":"post","link":"https:\/\/jjeci.net\/index.php\/2025\/03\/26\/global-modeling-and-predicting-of-thermal-conductivity-of-nanofluids-using-an-ensemble-of-different-tree-based-gradient-boosting-algorithms\/","title":{"rendered":"Global Modeling and Predicting of Thermal Conductivity of Nanofluids Using an Ensemble of Different Tree-based Gradient Boosting Algorithms"},"content":{"rendered":"\n

Authors: Bayan Ali Alyousef <\/strong>, Enas W. Abdulhay <\/strong>,  Ruba E. Khnouf <\/strong>,<\/strong><\/p>\n\n\n\n

DOI:\u00a0https:\/\/doi.org\/10.48103\/jjeci822025<\/a><\/p>\n\n\n\n

JORDANIAN JOURNAL OF ENGINEERING AND CHEMICAL INDUSTRIES (JJECI)<\/a><\/p>\n\n\n\n

Pages: 9-27<\/p>\n\n\n\n

\"\"<\/a><\/figure>\n\n\n\n
<\/div>\n\n\n

Abstract<\/h2>\n

\"\"<\/p>\n

Accurate prediction of the thermal conductivity of nanofluids attracts great interest from scholars, especially with experimental studies being so laborious and expensive, and with the inability of theoretical\/ empirical models to achieve the required accuracy. In addition, there is not enough
literature targeting the global modeling of thermal conductivity, in other words, the modeling of the mega profile -over various nanofluids- not of the thermal conductivity of a specific nanofluid. In this research, Extreme Gradient Boosting (XGBOOST), Light Gradient Boosting Machine
(LGBM), and Multilayer Perceptron (MLP) are implemented and optimized to predict nanofluids’ thermal conductivity in a global manner. Therefore, eight parameters; the temperature of nanofluids, size of nanoparticles, nanoparticles\u2019 volume concentration, thermal conductivity of the base fluid, thermal conductivity of nanoparticles, nanoparticle density, specific surface area of nanoparticles, and the nanoparticles\u2019 shape are chosen as model input variables. 4689 data points -representing various nanofluids and collected from 88 published papers – have been used to train the three mentioned models. Moreover, ten different sub-sets of features were investigated to detect the most important sub-set. The results from the three models as well as from the different subsets were then compared. Furthermore, the feature importance was determined for XGBOOST and LGBM. The results demonstrate a new approach to predict the global thermal conductivity. The best RMSE value on the
validation set is 0.0052 for MLP model, 0.011035 for LGBM model, and 0.00695 for the XGBOOST model. Also, the best sub-set includes size, temperature, concentration, thermal conductivity of base fluid, density, thermal conductivity of nanoparticle and shape. In addition, the highest relative importance of the nanofluids thermal conductivity is the thermal conductivity of base fluid. The shape is rarely explored in global prediction of thermal conductivity in literature.<\/p>\n

Paper type: Research paper<\/p>\n

Keywords: Nanofluids, Thermal Conductivity, MLP, XGBOOST, LGBM.<\/p>\n

Citation: Alyousef, B., Abdulhay, E., and Khnouf, R. .\u201c Global Modeling and Predicting of Thermal Conductivity of Nanofluids Using an Ensemble of Different Tree-based Gradient Boosting Algorithms\u201d, Jordanian Journal of Engineering and Chemical Industries,
Vol. 8, No.1, pp: 9-27 (2025).<\/p>\n\n\n

<\/div>\n\n\n\n
<\/a><\/div><\/div>\n","protected":false},"excerpt":{"rendered":"

Authors: Bayan Ali Alyousef , Enas W. Abdulhay ,  Ruba E. Khnouf , DOI:\u00a0https:\/\/doi.org\/10.48103\/jjeci822025 JORDANIAN JOURNAL OF ENGINEERING AND CHEMICAL INDUSTRIES (JJECI) Pages: 9-27 Abstract Accurate prediction of the thermal conductivity of nanofluids attracts great interest from scholars, especially with experimental studies being so laborious and expensive, and with the inability of theoretical\/ empirical models …<\/p>\n","protected":false},"author":1,"featured_media":1687,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[66],"tags":[],"class_list":{"0":"post-1645","1":"post","2":"type-post","3":"status-publish","4":"format-standard","5":"has-post-thumbnail","7":"category-volume-8-issue-1"},"_links":{"self":[{"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/posts\/1645","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/comments?post=1645"}],"version-history":[{"count":8,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/posts\/1645\/revisions"}],"predecessor-version":[{"id":1823,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/posts\/1645\/revisions\/1823"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/media\/1687"}],"wp:attachment":[{"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/media?parent=1645"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/categories?post=1645"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/jjeci.net\/index.php\/wp-json\/wp\/v2\/tags?post=1645"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}