Interactive Bayesian optimization : learning user preferences for graphics and animation

Brochu, Eric

doi:10.14288/1.0051462

Cited by 4 publications

(1 citation statement)

References 64 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The Gaussian process has desirable properties such as uncertainty estimates over function values, resistance to over tting, and principled approaches to hyperparameter optimization (Gal, 2015). The Gaussian process is e cient and applied when very little information about the objective function is available, making it useful for optimizing costly black-box functions (Brochu, 2010). Gaussian process speci ed by its mean function m(x) and covariance function k x, x * .…”

Section: Theoretical Background Of Bayesian Optimization Techniquementioning

confidence: 99%

Implementation of Bayesian Hyperparameter Optimization for Predicting Student Dropout in Sub-Saharan Africa Secondary Schools

Mnyawami

Maziku

Mushi

2022

Preprint

View full text Add to dashboard Cite

Many secondary schools in Sub-Saharan countries face the problem of students dropping out of school due to various reasons which are difficult to diagnose directly. Various initiatives such as the big results now initiatives, free education for all, no child left behind, and secondary education development programme as well as machine learning prediction models used to reduce the severity of the problem in Sub-Saharan countries. The ongoing dropout problem, particularly in secondary schools, is ascribed to improper root cause identification and the absence of formal procedures that may be used to estimate the severity of the issue. Bayesian Optimization technique has been used to project the student dropout by optimizing the prediction results. The forecast accuracy of machine learning algorithms is hampered by default optimization techniques, making it difficult to pinpoint the real causes of student dropouts. The performance of the optimized model was evaluated by the average accuracy, precision, recall, f1, Roc curve, and AUC. Prediction accuracy results indicate that NB = 95%, LDA = 94%, SGD = 96%, DT = 97%, RF = 97%, LR = 96%, KNN = 97%, and AdaBoost = 97%. Features including student marks, age, school distance, parental education, number of children, and parental occupation significantly contribute to student dropouts. Results indicate the prediction accuracy of the proposed model outperforms the default optimization method of the machine learning algorithms. A well-optimized strategy draws a lot of attention to the findings related to student dropout rates in Sub-Saharan countries.

show abstract

Section: Theoretical Background Of Bayesian Optimization Techniquementioning

confidence: 99%