How false data affects machine learning models in electrochemistry?
Source abstract
Abstract
Recently, the selection of machine learning model based on only the data distribution without concerning the noise of the data. This study aims to distinguish, which models perform well under noisy data, and establish whether stacking machine learning models actually provide robustness to otherwise weak-to-noise models. The electrochemical data were tested with 12 standalone models and stacking model. This includes XGB, LGBM, RF, GB, ADA, NN, ELAS, LASS, RIDGE, SVM, KNN, DT, and the stacking model. It is found that linear models handle noise well with the average error of (slope) to 1.75 F g-1 up to error per 100% percent noise added; but it suffers from prediction accuracy due to having an average of 60.19 F g-1 estimated at minimal error at 0% noise added. Tree-based models fail in terms of noise handling (average slope is 55.24 F g-1 at 100% percent noise), but it can provide higher prediction accuracy (lowest error of 23.9 F g-1) than that of linear. To address the controversial between prediction accuracy and error handling, the stacking model was constructed, which is not only show high accuracy (intercept of 25.03 F g-1), but it also exhibits good noise handling (slope of 43.58 F g-1), making stacking models a relatively low risk and viable choice for beginner and experienced machine learning research in electrochemistry. Even though neural networks (NN) are gaining popularity in the electrochemistry field. However, this study presents that NN is not suitable for electrochemical data, and improper tuning resulting in a model that is susceptible to noise. Thus, STACK models should provide better benefits in that even with untuned base models, they can achieve an accurate and noise-tolerant model. Overall, this work provides insight into machine learning model selection for electrochemical data, which should aid the understanding of data science in chemistry context.
Source: openaire · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Machine Learning in Materials Science · Computational Drug Discovery Methods · Electrochemical Analysis and Applications
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (54)
Show source-provided references
1. Is machine learning overhyped?10.1021/cen-09634-feature1 ↗
2. Embedded atom neural network potentials: efficient and accurate machine learning with a physically inspired representation10.1021/acs.jpclett.9b02037 ↗
3. Estimation of unit process data for life cycle assessment using a decision tree-based approach10.1021/acs.est.0c07484 ↗
4. Functional output regression for machine learning in materials science10.1021/acs.jcim.2c00626 ↗
5. Zhao · 2022
6. ChatGPT chemistry assistant for text mining and the prediction of MOF synthesis10.1021/jacs.3c05819 ↗
7. Insights into heteroatom-doped graphene supercapacitor data through manual data separation and statistical analysis
8. Enhancing data analysis with noise removal10.1109/TKDE.2006.46 ↗
9. Preventing human error: the impact of data entry methods on data accuracy and statistical results10.1016/j.chb.2011.04.004 ↗
10. Reducing quantitative data errors: tips for clinical researchers
11. Errors within the total laboratory testing process, from test selection to medical decision-making–A review of causes, consequences, surveillance and solutions10.11613/BM.2020.020502 ↗
12. Dealing with noise problem in machine learning data-sets: a systematic review10.1016/j.procs.2019.11.146 ↗
13. Budach · 2022
14. Data-driven machine learning approach for predicting the capacitance of graphene-based supercapacitor electrodes10.1016/j.est.2022.105411 ↗
15. Machine learning-based prediction of supercapacitor performance for a novel electrode material: cerium oxynitride10.1016/j.ensm.2021.05.024 ↗
16. The impact of physicochemical features of carbon electrodes on the capacitive performance of supercapacitors: a machine learning approach10.1038/s41598-023-33524-1 ↗
17. Aleatoric and epistemic uncertainty in machine learning: an introduction to concepts and methods10.1007/s10994-021-05946-3 ↗
18. A survey on epistemic (model) uncertainty in supervised learning: recent advances and applications10.1016/j.neucom.2021.10.119 ↗
19. Chemistry-informed machine learning prediction of compressive strength for alkali-activated materials10.1016/j.conbuildmat.2021.126103 ↗
20. Chemical signatures to identify the origin of solid ashes for efficient recycling using machine learning10.1016/j.jclepro.2022.133020 ↗
21. A user's guide to machine learning for polymeric biomaterials10.1021/acspolymersau.2c00037 ↗
22. Unlocking the full potential of heteroatom-doped graphene-based supercapacitors through stacking models and SHAP-guided optimization10.1021/acs.jcim.3c00670 ↗
23. In an overview of overfitting and its solutions
24. Noise detection and elimination in data preprocessing: experiments in medical domains10.1080/088395100117124 ↗
25. Noise reduction for instance-based learning with a local maximal margin approach10.1007/s10844-009-0101-z ↗
26. In Dealing with noise in defect prediction
27. Kordos · 2015 · 73
28. Noise reduction for instance-based learning with a local maximal margin approach10.1007/s10844-009-0101-z ↗
29. Random forests10.1023/A:1010933404324 ↗
30. A short introduction to boosting
31. The impact of simulated spectral noise on random forest and oblique random forest classification performance10.1155/2018/8316918 ↗
32. A study of effects of multicollinearity in the multivariable analysis
33. Mitigating the multicollinearity problem and its machine learning approach: a review10.3390/math10081283 ↗
34. Chowdhury · 2022 · 17
35. Multicollinearity effect in regression analysis: a feed forward artificial neural network approach10.9734/ajpas/2020/v6i130151 ↗
36. Enhancing graphene capacitance by nitrogen: effects of doping configuration and concentration10.1039/C5CP06952A ↗
37. Understanding the impact of nitrogen doping and/or amine functionalization of reduced graphene oxide via hydrothermal routes for supercapacitor applications10.1016/j.electacta.2021.139241 ↗
38. Nitrogen doping of graphene and its effect on quantum capacitance, and a new insight on the enhanced capacitance of N-doped carbon10.1039/c2ee23442d ↗
39. Biosourced quinones for high-performance environmentally benign electrochemical capacitors via interface engineering10.1038/s42004-022-00719-y ↗
40. Oxygen functional groups improve the energy storage performances of graphene electrochemical supercapacitors10.1039/C7RA12425B ↗
41. Machine learning approach to understanding the ‘synergistic’pseudocapacitive effects of heteroatom doped graphene10.1088/2053-1583/acaf8d ↗
42. 2-aminoanthraquinone anchored on N-doped reduced graphene oxide for symmetric supercapacitor with boosting energy density10.1016/j.electacta.2023.142194 ↗
43. Synergistic effect of oxygen and nitrogen Co-doping in metal–organic framework-derived ultramicroporous carbon for an exceptionally stable solid-state supercapacitor via a “proton trap” mechanism10.1021/acs.energyfuels.1c00918 ↗
44. Insights from machine learning of carbon electrodes for electric double layer capacitors10.1016/j.carbon.2019.08.090 ↗
45. Lundberg · 2017
46. In the relationship of surface area to cell capacitance for monolith carbon electrode from biomass materials for supercapacitor aplication
47. Controlling the effective surface area and pore size distribution of sp2 carbon materials and their impact on the capacitance performance of these materials10.1021/ja402552h ↗
48. Capacitive dominated charge storage in supermicropores of self-activated carbon electrodes for symmetric supercapacitors10.1016/j.est.2022.104776 ↗
49. New perspectives on the charging mechanisms of supercapacitors10.1021/jacs.6b02115 ↗
50. Carbonized wood for supercapacitor electrodes10.1149/2.005405ssl ↗
51. Evaluation of physical and electrochemical performances of hardwood and softwood derived activated carbons for supercapacitor application10.1016/j.mset.2022.09.002 ↗
52. High-energy-density flexible graphene-based supercapacitors enabled by atypical hydroquinone dimethyl ether
53. Thermal treatment effects on charge storage performance of graphene-based materials for supercapacitors10.1021/am300593k ↗
54. How many scientists fabricate and falsify research? A systematic review and meta-analysis of survey data10.1371/journal.pone.0005738 ↗