Predicting adhesion strength of micropatterned surfaces using gradient boosting models and explainable artificial intelligence visualizations

Ekanayake, I.U.; Palitha, Sandini; Gamage, Sajani; Meddage, D. P. P.; Wijesooriya, K.; Mohotti, Damith

doi:10.1016/j.mtcomm.2023.106545

Cited by 11 publications

(4 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Gradient Boosting (GB) is an algorithm that creates an ensemble of decision trees progressively to obtain precise predictions [23]. It begins from an initially poor model whose first tree in the ensemble is shallow and later adds additional trees.…”

Section: Gradient Boosting Regressormentioning

confidence: 99%

See 1 more Smart Citation

Optimizing Marketing Strategies: Integration of Al-Biruni Earth Radius Algorithm for Feature Selection and Pipeline Regression Model

Gaber,

Zaki,

Eid

et al. 2024

Journal of Artificial Intelligence in Engineering Practice

View full text Add to dashboard Cite

With the current business environment becoming increasingly ferocious, the effectiveness of digital marketing strategies is no longer a matter of debate as many organizations have realized the need to gain an edge over competition and improve the ROI with their marketing efforts. This study looks into the specifics of digital marketing effectiveness by, in the process, analyzing true indicators and key metrics. Demonstrating an understanding of the complexity of online marketing operations and the diversity of the variables involved, econometric techniques provide feature choice that affects campaign outcomes the most. At first, the variety of performance between different algorithms from feature selection gave the average error ranging from 0.38264 to 0.44194. However, following the optimization provides the tendency to see a decrease in mean errors and an improving performance. Afterward, the step of predictive modeling is implemented, employing diverse machine learning algorithms including ExtraTreesRegressor, GradientBoostingRegressor, SVR, and CatBoost to assess the effectiveness of foreshowing marketing outcomes. Before the optimization, the recommendations made by the predictive modeling are not too accurate and uniform for each algorithm. That being said, however, once the optimization is done, enhancement in prediction accuracy to the tune of substantial improvement is observed with metrics indicating the same as less MSE, RMSE, and R2. Contributing to a more thorough comprehension of the issue of selecting features and models for predicting as well as efficiency of digital marketing, the study also offers an understanding of the opportunities and obstacles that are present in the process of building digital marketing strategies. A thorough evaluation of top metrics and KPIs gives decision-makers data-driven tools to define their marketing activities, deliver tangible results, and stay relevant in the fast-paced digital environment of today.

show abstract

Section: Gradient Boosting Regressormentioning

confidence: 99%

“…XGB is a highly popular ML algorithm that performs very well in many application areas [23]. It gradually constitutes an assembly of trees, with each tree learning from the mistakes of the preceding ones.…”

Section: Xgboostmentioning

confidence: 99%

Optimizing Marketing Strategies: Integration of Al-Biruni Earth Radius Algorithm for Feature Selection and Pipeline Regression Model

Gaber,

Zaki,

Eid

et al. 2024

Journal of Artificial Intelligence in Engineering Practice

View full text Add to dashboard Cite

show abstract

“…In order to get more interpretable models, the development of explainable ensemble techniques will remain a priority. These models will elucidate the contributions of individual models within the ensemble, addressing regulatory compliance and facilitating decision-making in drug development [93,94].…”

Section: Future Directions In Ensemble Qsar Modelingmentioning

confidence: 99%

Ensemble Machine Learning Approach for Quantitative Structure Activity Relationship Based Drug Discovery: A Review

Noviandy,

Maulana,

Idroes

et al. 2023

Infolitika J. Data Sci.

View full text Add to dashboard Cite

This comprehensive review explores the pivotal role of ensemble machine learning techniques in Quantitative Structure-Activity Relationship (QSAR) modeling for drug discovery. It emphasizes the significance of accurate QSAR models in streamlining candidate compound selection and highlights how ensemble methods, including AdaBoost, Gradient Boosting, Random Forest, Extra Trees, XGBoost, LightGBM, and CatBoost, effectively address challenges such as overfitting and noisy data. The review presents recent applications of ensemble learning in both classification and regression tasks within QSAR, showcasing the exceptional predictive accuracy of these techniques across diverse datasets and target properties. It also discusses the key challenges and considerations in ensemble QSAR modeling, including data quality, model selection, computational resources, and overfitting. The review outlines future directions in ensemble QSAR modeling, including the integration of multi-modal data, explainability, handling imbalanced data, automation, and personalized medicine applications while emphasizing the need for ethical and regulatory guidelines in this evolving field.

show abstract

“…Explainable Artificial Intelligence (XAI) seeks to address the previous limitation of insufficient model interpretation 28 . XAI provides the reasoning behind how a particular prediction is made 29 . Hence, XAI is very popular in the context that dealing with high stakes 30 .…”

Section: Introductionmentioning

confidence: 99%

Modeling strength characteristics of basalt fiber reinforced concrete using multiple explainable machine learning with a graphical user interface

Kulasooriya,

Ranasinghe,

Perera

et al. 2023

Sci Rep

View full text Add to dashboard Cite

This study investigated the importance of applying explainable artificial intelligence (XAI) on different machine learning (ML) models developed to predict the strength characteristics of basalt-fiber reinforced concrete (BFRC). Even though ML is widely adopted in strength prediction in concrete, the black-box nature of predictions hinders the interpretation of results. Among several attempts to overcome this limitation by using explainable AI, researchers have employed only a single explanation method. In this study, we used three tree-based ML models (Decision tree, Gradient Boosting tree, and Light Gradient Boosting Machine) to predict the mechanical strength characteristics (compressive strength, flexural strength, and tensile strength) of basal fiber reinforced concrete (BFRC). For the first time, we employed two explanation methods (Shapley additive explanations (SHAP) and local interpretable model-agnostic explanations (LIME)) to provide explanations for all models. These explainable methods reveal the underlying decision-making criteria of complex machine learning models, improving the end user's trust. The comparison highlights that tree-based models obtained good accuracy in predicting strength characteristics yet, their explanations were different either by the magnitude of feature importance or the order of importance. This disagreement pushes towards complicated decision-making based on ML predictions which further stresses (1) extending XAI-based research in concrete strength predictions, and (2) involving domain experts to evaluate XAI results. The study concludes with the development of a “user-friendly computer application” which enables quick strength prediction of basalt fiber reinforced concrete (BFRC).

show abstract

Predicting adhesion strength of micropatterned surfaces using gradient boosting models and explainable artificial intelligence visualizations

Cited by 11 publications

References 44 publications

Optimizing Marketing Strategies: Integration of Al-Biruni Earth Radius Algorithm for Feature Selection and Pipeline Regression Model

Optimizing Marketing Strategies: Integration of Al-Biruni Earth Radius Algorithm for Feature Selection and Pipeline Regression Model

Ensemble Machine Learning Approach for Quantitative Structure Activity Relationship Based Drug Discovery: A Review

Modeling strength characteristics of basalt fiber reinforced concrete using multiple explainable machine learning with a graphical user interface

Contact Info

Product

Resources

About