Soil Compaction Parameters Prediction Based on Automated Machine Learning Approach

Abdullah Hulusi Kokcam; Alparslan Serhat Demir; Caner Erden; Talas Fikret Kurnaz; Ugur Dagdeviren

REVIEW 3 major objections 6 minor 21 references

Soil Compaction Parameters Prediction Based on Automated Machine Learning Approach

T0 review · 3 major / 6 minor · reviewed 2026-08-03 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The paper claims that an automated machine learning pipeline, selecting and tuning its own models, can predict optimum moisture content and maximum dry density from five routine soil index properties, with reported test R² values of 89.1% f

desk verdict A modest AutoML benchmark on a published 115-sample compaction dataset; plausible internal scores, but the headline generalization claim is unsupported because the cited external validation results never appear. read the letter →

arxiv 2512.08343 v1 pith:2KZMRPMU submitted 2025-12-09 cs.AI

Caner Erden , Alparslan Serhat Demir , Abdullah Hulusi Kokcam , Talas Fikret Kurnaz , Ugur Dagdeviren This is my paper

classification cs.AI

keywords soilcompactionoptimummoisturecontentmaximumdrydensityautomatedmachinelearningXGBoostfeatureimportanceProctortesttabularregression

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that an automated machine learning approach, which selects and tunes models without manual intervention, can predict two soil compaction parameters—the optimum moisture content (OMC) and the maximum dry density (MDD)—from five routinely measured soil index properties: liquid limit, plastic limit, gravel, sand, and fines percentages. On a heterogeneous 115-sample dataset spanning nine soil classes, the best configuration produced R² values of about 89.1% for OMC and 80.4% for MDD on a held-out test set. If correct, construction engineers could estimate compaction parameters from cheap index tests rather than time-consuming Proctor tests, and the approach could scale to new soil types without hand-tuning models. The paper also argues that the automated process, rather than any single algorithm, provides the robustness, and that heterogeneous training data is what enables generalization across soil types.

What carries the argument

The central mechanism is the automated machine learning pipeline itself: an off-the-shelf tabular prediction library that takes the soil dataset, tries a suite of tree-based, nearest-neighbor, and neural models under four configurations (default, a high-quality preset, a hyperparameter variant, and both), and returns a weighted ensemble. The heavy lifting is done by the pipeline's automated model selection and hyperparameter optimization rather than by a hand-built architecture. The named algorithm that carries the final result is Extreme Gradient Boosting (XGBoost), a gradient-boosted decision tree method, which under the best configuration achieved the reported test R² values. The feature

What would settle it

A reader could take the same five inputs on an independent multi-class soil dataset, fix the exact XGBoost configuration the paper names as best, and use a defined holdout or 5-fold split; if R² drops well below 0.80/0.89 or the cross-validation spread is wide, the headline values are artifacts of split choice. Publishing the results of the three external datasets cited for validation would settle the generalization claim directly.

Watch

Extended reading notes

Core claim

The paper's central claim is that off-the-shelf automated machine learning can serve as a reliable predictor of soil compaction parameters across differing soil types. Using only liquid limit, plastic limit, gravel, sand, and fines percentages as inputs, the trained models reproduce optimum moisture content with R² = 0.891 and maximum dry density with R² = 0.804 on a separate test portion of the primary dataset. The winning model in both cases was Extreme Gradient Boosting (XGBoost), a gradient-boosted decision tree algorithm, selected automatically under the most aggressive configuration. The paper argues this demonstrates that AutoML's automated algorithm selection and hyperparameter optim

Load-bearing premise

The generalization claim depends on the assumption that the held-out test set is representative and uncontaminated by model selection, and that the cited but unreported external evaluations actually confirmed the model; if either is false, the reported R² values do not show what the conclusion claims.

Editorial extensions

If this is right

If the result holds, engineers can screen compaction parameters from routine index tests before deciding whether full Proctor testing is needed, cutting laboratory time and cost.
Because the pipeline retrains automatically, updating the model with new site-specific data does not require expert machine-learning tuning.
The approach supports integration with smart construction systems: sensor data from compaction equipment could feed the same pipeline for real-time quality control and automated equipment adjustments.
The feature importance rankings give a field signal that liquid limit, plastic limit, and fines content matter most, while gravel content contributes little.
Reported success on a nine-class dataset suggests the method may extend to other soil classification systems, provided the training data remain heterogeneous.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The reported R² values are likely optimistic because the best model was selected by test-set score across many configurations and the train/test split details are not disclosed; a fair estimate would require nested cross-validation or a fixed split chosen before model selection.
The three external datasets cited for validation are named but their results never appear in the manuscript; without those numbers, the 'across different soil types' conclusion rests on a single 115-sample dataset.
A practical, testable deployment would be a conservative decision rule: use the AutoML prediction as a preliminary estimate and trigger laboratory confirmation only when predicted OMC or MDD falls outside a safe band—an extension the paper does not spell out.
Since all inputs are standard index properties, the method could readily be packaged as a spreadsheet or web calculator for preliminary geotechnical estimates, a natural but unstated consequence.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Short version: this is a workmanlike AutoML benchmark applied to a published soil compaction dataset. The application is new to this problem area, and the paper is readable and transparent about the tooling. But the headline generalization claim is unsupported: Section 2 promises validation on three external datasets [18–20], and the rest of the paper reports only test-set results from Günaydın (2009). If those external runs produced poor numbers, omitting them hides a failed check; if they were never run, the claim is simply unjustified. The abstract's phrase 'on a separate dataset' does not fix the ambiguity.

What the paper does well: it is the first AutoML application for compaction parameters I know of, it uses a heterogeneous set of 115 samples from nine soil classes, removes duplicates, and reports full tables of model scores rather than cherry-picking one result. Feature importance with permutation-style stats is a plus, and LL coming out dominant is consistent with prior work. The internal test scores are plausible for this kind of data.

The soft spots are real but not fatal. The best model is selected by test score across dozens of model/config combinations, so the reported R² is optimistic; there is no fixed split description, no repeated seeds or confidence intervals. The paper never compares against the earlier ML baselines on the same data (e.g., the ANN/regression results in Günaydın 2009 or Kurnaz & Kaya 2020), which makes it hard to know what AutoML actually adds. And the missing external validation is the load-bearing gap: 'robustness across diverse soil types' requires exactly the numbers that are absent.

I would not desk-reject this. A serious referee could send it back for a much shorter revision: report the external validation results or remove the claim, fix the evaluation protocol with a single holdout or cross-validation and error bars, and add a direct comparison to the published ML baselines. If the external validation turns out weak, the paper still has a place as a benchmark with caveats.

For me: I would not cite it until the external numbers appear. I would bring it to a reading group only if someone is specifically interested in AutoML practice; otherwise it is too narrow.

Referee Report

3 major / 6 minor

Summary. The paper proposes using the AutoGluon AutoML library to predict optimum moisture content (OMC) and maximum dry density (MDD) from liquid limit, plastic limit, gravel, sand, and fines content. The authors use 115 samples of nine soil classes from Günaydın (2009), apply four AutoGluon configurations, and report tables of R² values for many models. Headline results are XGBoost test R² of 0.891 for OMC (Table 5) and 0.804 for MDD (Table 8), claimed to demonstrate AutoML's effectiveness across different soil types. The manuscript states that external validation was performed on three additional datasets [18–20], but no such results appear anywhere in the paper.

Significance. If the conclusions were fully supported, the paper would offer a useful demonstration that off-the-shelf AutoML can predict soil compaction parameters from index properties on a heterogeneous dataset, potentially reducing laboratory effort. The paper's strengths include a transparent comparison of many model architectures and configurations on a publicly available dataset, and an explicit focus on the value of heterogeneous data. The main weakness is that the evidence as presented does not substantiate the generalization claim: the only reported test results come from a single source dataset, while the promised external validation is missing. The reported performance is also likely optimistically biased because the best model is selected by test score without accounting for multiple comparisons. These gaps are fixable but affect the central claims.

major comments (3)

[Section 2, Sections 4–5, Abstract, Conclusion] Section 2 states that 'validation tests were carried out on different datasets in the literature' and names three datasets [18–20], but Sections 4 and 5 report only test results from Günaydın (2009). The abstract and conclusion claim effectiveness and robustness 'across different soil types' based on this missing validation. If the external validations were performed, their results must be reported; if they were not performed, the claim is unsupported. Either way, the manuscript as written is internally inconsistent on this load-bearing point.
[Tables 5 and 8, Abstract] The headline R² values are selected as the best test scores over many model/configuration combinations: 0.891 for OMC from XGBoost_BAG_L1_FULL (Table 5) and 0.804 for MDD from XGBoost_BAG_L1 (Table 8). The best model is chosen using the test set, so the reported numbers are optimistically biased. No repeated seeds, cross-validation, or confidence intervals are provided. The authors should either use a proper nested protocol or report the distribution of scores across repeated splits and clarify exactly what 'Test Score' and 'Validation Score' represent in AutoGluon.
[Introduction, Section 4] The paper motivates AutoML as a way to overcome limitations of traditional ML models and cites prior ANN/SVR/GP studies, but it reports no comparison with any published model on the same data. To establish that AutoML 'improves prediction accuracy and generalizability,' the authors should compare with at least one baseline trained on the same Günaydın dataset (e.g., the ANN in [4] or the soft-computing methods in [8]). Without such a comparison, the claimed advantage over prior work is not demonstrated.

minor comments (6)

[Table 10] The caption contains 'Hata! Yer işareti başvurusu geçersiz.' (an unresolved cross-reference error) and should be fixed. The feature importance method (permutation? SHAP?) is not described; please specify.
[Abstract] The phrase 'on a separate dataset' is ambiguous. It likely refers to the internal test split from Günaydın (2009), but could be read as an external dataset. Please clarify.
[Section 2] The train/validation/test split ratio is not disclosed. State how the 115 samples were divided and whether stratification by soil class was used.
[Section 4, Tables 2–10] The tables report training time and scores but not error metrics such as RMSE or MAE. For engineering applications, reporting these in the same units would strengthen the practical interpretation.
[Section 3] The role of the 'multimodal' hyperparameter is unclear; a brief technical explanation or reference would help reproducibility.
[References] References [18]–[20] are theses; consider giving English titles and available identifiers for findability, and ensure the notation 'Günaydın (2009)' is used consistently as 'Günaydın [4]'.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: empirical AutoML benchmark reports genuine held-out predictions; reported gaps (missing external validation, model selection on test scores) are validity concerns, not constructional circularity.

full rationale

This paper contains no derivation chain that could be circular: it is an empirical AutoML benchmark in which AutoGluon predicts OMC and MDD from index properties (LL, PL, G%, S%, F%) and reports R² values on a held-out test set. No fitted parameter is renamed as a prediction, no output is used to define an input, and no uniqueness theorem or ansatz is imported from the authors' prior work. The self-citations, including [16] by overlapping authors, are used only to situate prior AutoML applications and are not load-bearing for the compaction prediction claim. The main concerns are methodological rather than circular: the paper states that validation tests were carried out on three external datasets [18–20] but never reports their results, and the headline XGBoost R² values are selected as the best test scores across many models and configurations in Tables 2–10, which makes the reported performance optimistic. These are evidence-reporting and model-selection issues, not instances where the result is equivalent to its input by construction. Under the stated rubric, no circular step can be exhibited, so the circularity score is 0.

Assumptions & free parameters 3 free parameters · 3 assumptions · 0 invented entities

The central claim has few invented entities or fitted constants, but it depends on domain assumptions about input sufficiency and dataset representativeness, plus an undisclosed split and test-set-based model selection. The AutoGluon hyperparameters are auto-fit; final values are not reported.

free parameters (3)

AutoGluon configuration/preset choices = 4 configs: default, Best Quality, multimodal, Best Quality+multimodal
Hand-chosen experimental designs; the best test score among them is reported, inflating headline R².
Train/validation/test split (undisclosed) = Not reported
The split method/ratio/seed is not given; all reported R² depend on it.
Hyperparameters tuned by AutoGluon = Auto-selected, not reported
Automated search is the method, but final hyperparameters are not disclosed, so exact reproduction is impossible.

assumptions (3)

domain assumption Five index properties (LL, PL, G%, S%, F%) are sufficient inputs to predict OMC and MDD across nine soil classes.
Inherited from Günaydın [4]; the paper does not test other variables such as compaction energy or categorical soil class.
domain assumption The 115-sample dataset is representative of diverse soil types.
The paper relies on nine classes but only 115 samples; external validation is not reported.
ad hoc to paper AutoGluon's automated model selection and test-set-based choice of the best configuration gives an unbiased performance estimate.
The paper selects the best model by test score across many configurations, which violates unbiasedness and biases the headline R².

how reviews work

0 comments

Cite this review

Pith. "Pith review of Soil Compaction Parameters Prediction Based on Automated Machine Learning Approach." pith.science (2026). https://pith.science/paper/2KZMRPMU

@misc{pith2026251208343,
  author       = {Pith},
  title        = {Pith review of: Soil Compaction Parameters Prediction Based on Automated Machine Learning Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/2KZMRPMU}},
  note         = {Machine review of arXiv:2512.08343}
}

read the original abstract

Soil compaction is critical in construction engineering to ensure the stability of structures like road embankments and earth dams. Traditional methods for determining optimum moisture content (OMC) and maximum dry density (MDD) involve labor-intensive laboratory experiments, and empirical regression models have limited applicability and accuracy across diverse soil types. In recent years, artificial intelligence (AI) and machine learning (ML) techniques have emerged as alternatives for predicting these compaction parameters. However, ML models often struggle with prediction accuracy and generalizability, particularly with heterogeneous datasets representing various soil types. This study proposes an automated machine learning (AutoML) approach to predict OMC and MDD. AutoML automates algorithm selection and hyperparameter optimization, potentially improving accuracy and scalability. Through extensive experimentation, the study found that the Extreme Gradient Boosting (XGBoost) algorithm provided the best performance, achieving R-squared values of 80.4% for MDD and 89.1% for OMC on a separate dataset. These results demonstrate the effectiveness of AutoML in predicting compaction parameters across different soil types. The study also highlights the importance of heterogeneous datasets in improving the generalization and performance of ML models. Ultimately, this research contributes to more efficient and reliable construction practices by enhancing the prediction of soil compaction parameters.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

21 extracted references · 12 canonical work pages

[4]

Estimation of soil compaction parameters by using statistical analyses and artificial neural networks,

O. Günaydın, “Estimation of soil compaction parameters by using statistical analyses and artificial neural networks,” Environ Geol , vol. 57, no. 1, pp. 203 –215, Mar. 2009, doi: 10.1007/s00254-008-1300-6

work page doi:10.1007/s00254-008-1300-6 2009
[8]

The performance comparison of the soft computing methods on the prediction of soil compaction parameters,

T. F. Kurnaz and Y. Kaya, “The performance comparison of the soft computing methods on the prediction of soil compaction parameters,” Arab J Geosci, vol. 13, no. 4, p. 159, Feb. 2020, doi: 10.1007/s12517-020-5171-9

work page doi:10.1007/s12517-020-5171-9 2020
[1]

Prediction of maximum dry density and optimum moisture content of stabilised soil using RBF neural networks,

A. H. Alavi, A. H. Gandomi, M. Gandomi, and S. S. Sadat Hosseini, “Prediction of maximum dry density and optimum moisture content of stabilised soil using RBF neural networks,” The IES Journal Part A: Civil & Structural Engineering , vol. 2, no. 2, pp. 98 –106, May 2009, doi: 10.1080/19373260802659226

work page doi:10.1080/19373260802659226 2009
[2]

Soil compaction parameters prediction using GMDH -type neural network and genetic algorithm,

A. Ardakani and A. Kordnaeij, “Soil compaction parameters prediction using GMDH -type neural network and genetic algorithm,” European Journal of Environmental and Civil Engineering, vol. 23, no. 4, pp. 449–462, Apr. 2019, doi: 10.1080/19648189.2017.1304269

arXiv 2019
[3]

Progressive Machine Learning Approaches for Predicting the Soil Compaction Parameters,

M. A. Benbouras and L. Lefilef, “Progressive Machine Learning Approaches for Predicting the Soil Compaction Parameters,” Transp. Infrastruct. Geotech., vol. 10, no. 2, pp. 211 –238, Apr. 2023, doi: 10.1007/s40515-021-00212-4

work page doi:10.1007/s40515-021-00212-4 2023
[5]

Estimating compaction parameters of fine - and coarse-grained soils by means of artificial neural networks,

F. Isik and G. Ozden, “Estimating compaction parameters of fine - and coarse-grained soils by means of artificial neural networks,” Environ Earth Sci , vol. 69, no. 7, pp. 2287 –2297, Aug. 2013, doi: 10.1007/s12665-012-2057-5

work page doi:10.1007/s12665-012-2057-5 2013
[6]

Predicting the compaction characteristics of expansive soils using two genetic programming -based algorithms,

F. E. Jalal, Y. Xu, M. Iqbal, B. Jamhiri, and M. F. Javed, “Predicting the compaction characteristics of expansive soils using two genetic programming -based algorithms,” Transportation Geotechnics, vol. 30, p. 100608, Sep. 2021, doi: 10.1016/j.trgeo.2021.100608

arXiv 2021
[7]

Prediction of compaction parameters of soils using artificial neural network,

J. Jayan and N. Sankar, “Prediction of compaction parameters of soils using artificial neural network,” Asian Journal of Engineering and Technology, vol. 3, no. 4, 2015

2015

Show all 21 references

[9]

Artificial Neural Network Prediction Models for Soil Compaction and Permeability,

S. K. Sinha and M. C. Wang, “Artificial Neural Network Prediction Models for Soil Compaction and Permeability,” Geotech Geol Eng , vol. 26, no. 1, pp. 47 –64, Feb. 2008, doi: 10.1007/s10706-007-9146-3

2008 doi
[10]

Prediction of Maximum Dry Density and Unconfined Compressive Strength of Cement Stabilised Soil Using Artificial Intelligence Techniques,

S. Suman, M. Mahamaya, and S. K. Das, “Prediction of Maximum Dry Density and Unconfined Compressive Strength of Cement Stabilised Soil Using Artificial Intelligence Techniques,” Int. J. of Geosynth. and Ground Eng., vol. 2, no. 2, p. 11, Jun. 2016, doi: 10.1007/s40891-016-0051...

2016 doi
[11]

Multi -layer perceptron (MLP) neural network for predicting the modified compaction parameters of coarse -grained and fine-grained soils,

G. Verma and B. Kumar, “Multi -layer perceptron (MLP) neural network for predicting the modified compaction parameters of coarse -grained and fine-grained soils,” Innov. Infrastruct. Solut., vol. 7, no. 1, p. 78, Feb. 2022, doi: 10.1007/s41062-021-00679-7

2022 doi
[12]

High performance prediction of soil compaction parameters using multi expression programming,

H.-L. Wang and Z.-Y. Yin, “High performance prediction of soil compaction parameters using multi expression programming,” Engineering Geology, vol. 276, p. 105758, Oct. 2020, doi: 10.1016/j.enggeo.2020.105758

2020
[13]

Greedy-AutoML: A novel greedy-based stacking ensemble learning framework for assessing soil liquefaction potential,

E. K. Sahin and S. Demir, “Greedy-AutoML: A novel greedy-based stacking ensemble learning framework for assessing soil liquefaction potential,” Engineering Applications of Artificial Intelligence, vol. 119, p. 105732, Mar. 2023, doi: 10.1016/j.engappai.2022.105732

2023
[14]

Landslide Susceptibility Assessment Using an AutoML Framework,

A. G. Bruzón et al. , “Landslide Susceptibility Assessment Using an AutoML Framework,” IJERPH, vol. 18, no. 20, p. 10971, Oct. 2021, doi: 10.3390/ijerph182010971

2021 doi
[15]

Prediction of Landslides and Determination of Its Variable Importance Using AutoML,

K. Nam, M. -I. Kim, O. Kwon, F. Wang, and G. -C. Jeong, “Prediction of Landslides and Determination of Its Variable Importance Using AutoML,” The Journal of Engineering Geology, vol. 30, no. 3, pp. 315–325, Eylül 2020, doi: 10.9720/KSEG.2020.3.315

2020 doi
[16]

Comparison of machine learning algorithms for slope stability prediction using an automated machine learning approach,

T. F. Kurnaz, C. Erden, U. Dağdeviren, A. S. Demir, and A. H. Kökçam, “Comparison of machine learning algorithms for slope stability prediction using an automated machine learning approach,” Nat Hazards, Mar. 2024, doi: 10.1007/s11069-024-06490-8

2024 doi
[17]

Machine Learning Models for Slope Stability Classification of Circular Mode Failure: An Updated Database and Automated Machine Learning (AutoML) Approach,

J. Ma et al., “Machine Learning Models for Slope Stability Classification of Circular Mode Failure: An Updated Database and Automated Machine Learning (AutoML) Approach,” Sensors, vol. 22, no. 23, p. 9166, Nov. 2022, doi: 10.3390/s22239166

2022 doi
[18]

Prediction of compaction characteristics of over -consolidated soils,

A. H. A. Hussain, “Prediction of compaction characteristics of over -consolidated soils,” PhD Thesis, PhD. Thesis, Near East University, 2016

2016
[19]

Yapay sinir ağları yaklaşımı ile kompaksiyon parametrelerinin tahmini,

Y. T. Soycan, “Yapay sinir ağları yaklaşımı ile kompaksiyon parametrelerinin tahmini,” PhD Thesis, Yüksek Lisans Tezi, Niğde Üniversitesi, Fen Bilimleri Enstitüsü, Niğde, 2008. Accessed: Feb. 22, 2024. [Online]. Available: https://platform.almanhal.com/Details/Thesis/200014761...

2008
[20]

Determination of compaction parameters by means of regression approaches,

A. Olmez, “Determination of compaction parameters by means of regression approaches,” Master’s thesis, 2007

2007
[21]

AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data,

N. Erickson et al., “AutoGluon-Tabular: Robust and Accurate AutoML for Structured Data,” Mar. 13, 2020, arXiv: arXiv:2003.06505. Accessed: Feb. 22, 2024. [Online]. Available: http://arxiv.org/abs/2003.06505 589

2020 arXiv

Pith tools

Reviewed August 3, 2026 · model on record in the stance chip above.

[1] [4]

Estimation of soil compaction parameters by using statistical analyses and artificial neural networks,

O. Günaydın, “Estimation of soil compaction parameters by using statistical analyses and artificial neural networks,” Environ Geol , vol. 57, no. 1, pp. 203 –215, Mar. 2009, doi: 10.1007/s00254-008-1300-6

work page doi:10.1007/s00254-008-1300-6 2009

[2] [8]

The performance comparison of the soft computing methods on the prediction of soil compaction parameters,

T. F. Kurnaz and Y. Kaya, “The performance comparison of the soft computing methods on the prediction of soil compaction parameters,” Arab J Geosci, vol. 13, no. 4, p. 159, Feb. 2020, doi: 10.1007/s12517-020-5171-9

work page doi:10.1007/s12517-020-5171-9 2020

[3] [1]

Prediction of maximum dry density and optimum moisture content of stabilised soil using RBF neural networks,

A. H. Alavi, A. H. Gandomi, M. Gandomi, and S. S. Sadat Hosseini, “Prediction of maximum dry density and optimum moisture content of stabilised soil using RBF neural networks,” The IES Journal Part A: Civil & Structural Engineering , vol. 2, no. 2, pp. 98 –106, May 2009, doi: 10.1080/19373260802659226

work page doi:10.1080/19373260802659226 2009

[4] [2]

Soil compaction parameters prediction using GMDH -type neural network and genetic algorithm,

A. Ardakani and A. Kordnaeij, “Soil compaction parameters prediction using GMDH -type neural network and genetic algorithm,” European Journal of Environmental and Civil Engineering, vol. 23, no. 4, pp. 449–462, Apr. 2019, doi: 10.1080/19648189.2017.1304269

arXiv 2019

[5] [3]

Progressive Machine Learning Approaches for Predicting the Soil Compaction Parameters,

M. A. Benbouras and L. Lefilef, “Progressive Machine Learning Approaches for Predicting the Soil Compaction Parameters,” Transp. Infrastruct. Geotech., vol. 10, no. 2, pp. 211 –238, Apr. 2023, doi: 10.1007/s40515-021-00212-4

work page doi:10.1007/s40515-021-00212-4 2023

[6] [5]

Estimating compaction parameters of fine - and coarse-grained soils by means of artificial neural networks,

F. Isik and G. Ozden, “Estimating compaction parameters of fine - and coarse-grained soils by means of artificial neural networks,” Environ Earth Sci , vol. 69, no. 7, pp. 2287 –2297, Aug. 2013, doi: 10.1007/s12665-012-2057-5

work page doi:10.1007/s12665-012-2057-5 2013

[7] [6]

Predicting the compaction characteristics of expansive soils using two genetic programming -based algorithms,

F. E. Jalal, Y. Xu, M. Iqbal, B. Jamhiri, and M. F. Javed, “Predicting the compaction characteristics of expansive soils using two genetic programming -based algorithms,” Transportation Geotechnics, vol. 30, p. 100608, Sep. 2021, doi: 10.1016/j.trgeo.2021.100608

arXiv 2021

[8] [7]

Prediction of compaction parameters of soils using artificial neural network,

J. Jayan and N. Sankar, “Prediction of compaction parameters of soils using artificial neural network,” Asian Journal of Engineering and Technology, vol. 3, no. 4, 2015

2015