The Rationale for the Use of Spectral Expansion Apparatus for Modelling the Electromechanical Systems

Burkov, Petr V.; Burkov, Vladimir P.; Timofeev, V.Yu.; Burkova, Svetlana P.

doi:10.26730/1816-4528-2019-1-54-62

Cited by 25 publications

(40 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The models presented in these studies used traditional machine learning algorithms (e.g., logistic regression, Naive Bayes, and support vector machines), which are very useful for simpler automated text analysis (ATA) tasks. 38 However, more complex writing and more nuanced labeling schemes may require more advanced machine learning methods, such as neural networks.…”

Section: Feedback In Chemistrymentioning

confidence: 99%

Automated Text Analysis of Organic Chemistry Students’ Written Hypotheses

Dood,

Watts,

Connor

et al. 2024

J. Chem. Educ.

View full text Add to dashboard Cite

Generating a testable hypothesis is a necessary skill for engaging in science, requiring both general reasoning skills and specific content knowledge of the phenomenon being investigated. While many students have the reasoning skills necessary for developing testable hypotheses in a general science context, it can be challenging for students to apply these skills to content areas where the phenomena being studied are unobservable, such as problems in organic chemistry. Generating hypotheses for experiments is a skill that students are expected to apply in chemistry teaching laboratories. Students can benefit from receiving real-time feedback to support developing this ability. However, providing feedback on writing can be challenging for instructors, especially in large-enrollment courses. For this study, we explored and compared the performance of several machine learning algorithms to classify organic chemistry students' written hypotheses for a science-general scenario and an organic chemistry-specific scenario. These models were trained to analyze students' written hypotheses for the presence of five features: (1) a testable prediction in a given situation; (2) a predicted change in the independent variable that is related to the dependent variable;(3) scientific content addressing the research question provided in the prompt; (4) a correct and completely defined independent variable; and (5) a correct and completely defined dependent variable. This work has implications for guiding future research on the feasibility of providing instructors with information about students' hypothesis-generating abilities that can support the delivery of real-time feedback.

show abstract

Section: Feedback In Chemistrymentioning

confidence: 99%

Automated Text Analysis of Organic Chemistry Students’ Written Hypotheses

Dood,

Watts,

Connor

et al. 2024

J. Chem. Educ.

View full text Add to dashboard Cite

show abstract

“…The authors state that, 'the best hyperparameters are those with a minimal MAE on the test set across the five folds.' Best practices in data science however dictate that, in order to minimize the risk of hyperparameter overfitting, one ought to only optimize hyperparameters with a validation set and use a held-out test set to accurately measure performance on unseen data [96,98,[172][173][174][175][176][177][178][179][180] i.e. if doing five-fold cross validation, it is essential to first create five separate training, validation, and testing splits.…”

Section: Modelsmentioning

confidence: 99%

Comment on ‘Physics-based representations for machine learning properties of chemical reactions’

Spiekermann,

Stuyver,

Pattanaik

et al. 2023

Mach. Learn.: Sci. Technol.

View full text Add to dashboard Cite

In a recent article in this journal, van Gerwen et al (2022 Mach. Learn.: Sci. Technol. 3 045005) presented a kernel ridge regression model to predict reaction barrier heights. Here, we comment on the utility of that model and present references and results that contradict several statements made in that article. Our primary interest is to offer a broader perspective by presenting three aspects that are essential for researchers to consider when creating models for chemical kinetics: (1) are the model’s prediction targets and associated errors sufficient for practical applications? (2) Does the model prioritize user-friendly inputs so it is practical for others to integrate into prediction workflows? (3) Does the analysis report performance on both interpolative and more challenging extrapolative data splits so users have a realistic idea of the likely errors in the model’s predictions?

show abstract

“…In a typical telecommunications system, typical concerns such as data, experimentation or model management, deployment, reproducibility, and testing & monitoring should be considered depending on the ML platform. In a ML project lifecycle management as described in [166], all business requirements and goals of the project must be defined first before the project starts. After the business requirements are co-decided and the project objectives are defined, data collection and preparation phase follows.…”

Section: A Gap Analysismentioning

confidence: 99%