Model selection of SVMs using GA approach

Chen, Peng-Wei; Wang, Jung-Ying; Lee, Hahn-Ming

doi:10.1109/ijcnn.2004.1380929

Cited by 30 publications

(10 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This allowed to identify 1694 2ν 6 −ν 5 and 210 3ν 6 −2ν 5 resonant asteroids. Machine learning models, optimized through the use of genetic algorithms [15], were created as a result of this analysis for both resonances. For the case of the 2ν 6 − ν 5 resonance, a deep learning CNN model was obtained and optimized to correct possible overfitting issues [9], which will allow the classifications of future populations of resonant asteroids in time-scales of seconds.…”

Section: Discussionmentioning

confidence: 99%

“…While the imbalance for both databases is not severe, with an imbalance ratio below 100, it is worth checking if methods for dealing with imbalanced data sets, as described in [42], may improve the performance of machine learning methods. We studied both data sets correcting the optimal methods identified by genetic algorithms [43] with approaches based on oversampling, undersampling, or both over-and undersampling of the minority classes. Our results are displayed in the supplementary material (1).…”

Section: Appendix a Supplementary Materials (1): Machine Learning Cla...mentioning

confidence: 99%

See 1 more Smart Citation

Deep learning classification of asteroids in g-type secular resonances

Carruba

Aljbaae

Knežević

et al. 2023

Preprint

View full text Add to dashboard Cite

Linear secular resonances happen when there is a commensurability between the precession frequency of the pericenter, g, or longitude of the node, s, of an asteroid and a planet. Non-linear resonances are higherorder combinations of these frequencies. Here, we studied the three most diffusive g−type non-linear secular resonances using Artificial Neural Networks. We identified a population of more than 2100 resonant objects in the g − 2g6 + g5 and g − 3g6 + 2g5 resonances. This allows the creation of a Convolutional Neural Network model for the g − 2g6 + g5 resonance, able to predict the status of several thousands of asteroids in seconds. We identified 12 new possible dynamical groups among the resonant population, including the 5507 and 170776 families, which have both estimated ages of less than 7 My. These are the two first-ever identified young families in resonant configurations of the investigated resonances, which allows setting limits on their original ejection velocity field.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Appendix a Supplementary Materials (1): Machine Learning Cla...mentioning

confidence: 99%

Deep learning classification of asteroids in g-type secular resonances

Carruba

Aljbaae

Knežević

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…Offering versatility, BaumEvA accommodates both binary and combinatorial data types, equipped with a comprehensive suite of optimization and search tools. It integrates various mechanisms for selection 66 , crossover 67 , mutation 68 , and parent selection 69 , ensuring robust performance across diverse optimization scenarios.…”

Section: Fitness Values Of Individualsmentioning

confidence: 99%

Two-Stage Genetic Algorhitm for Optimization Logistics Network for Groupage Delivery

Malashin,

Tynchenko,

Masich

et al. 2024

Preprint

View full text Add to dashboard Cite

This study explores groupage delivery optimization through a two-stage GA framework. The goal of optimization is to define locations of regional branch warehouses of Logistics Network to reduce transportation costs. The first stage of the GA focuses on identifying branch warehouses from a list of cities and uses the standard binary string approach, while the second stage solves Vehicle Routing Problem (VRP) using combinatorial GA. This research contributes to advancing logistics optimization methodologies and provides valuable insights for solving Location Routing Problem (LRP) which entails two problems: the Facility Location Problem (FLP) and the Multiple Vehicle Routing Problem (MVRP).

show abstract

“…No universally agreed method has been reported for the optimization of SVM parameters worldwide. At present, the common methods include test method, grid search (GS), genetic algorithm (GA), and particle swarm optimization (PSO) (Sánchez, 2003;Peng-Wei et al, 2004).…”

Section: Support Vector Machine (Svm)mentioning

confidence: 99%

Early Diagnosis of Type 2 Diabetes Based on Near-Infrared Spectroscopy Combined With Machine Learning and Aquaphotomics

Guo²,

et al. 2020

Front. Chem.

View full text Add to dashboard Cite

Early diagnosis is important to reduce the incidence and mortality rate of diabetes. The feasibility of early diagnosis of diabetes was studied via near-infrared spectra (NIRS) combined with a support vector machine (SVM) and aquaphotomics. Firstly, the NIRS of entire blood samples from the population of healthy, pre-diabetic, and diabetic patients were obtained. The spectral data of the entire spectra in the visible and near-infrared region (400–2,500 nm) were used as the research object of the qualitative analysis. Secondly, several preprocessing steps including multiple scattering correction, variable standardization, and first derivative and second derivative steps were performed and the best pretreatment method was selected. Finally, for the early diagnosis of diabetes, models were established using SVM. The first overtone of water (1,300–1,600 nm) was used as the research object for an aquaphotomics model, and the aquagram of the healthy group, pre-diabetes, and diabetes groups were drawn using 12 water absorption patterns for the early diagnosis of diabetes. The results of SVM showed that the highest accuracy was 97.22% and the specificity and sensitivity were 95.65 and 100%, respectively when the pretreatment method of the first derivative was used, and the best model parameters were c = 18.76 and g = 0.008583.The results of the aquaphotomics model showed clear differences in the 1,400–1,500 nm region, and the number of hydrogen bonds in water species (1,408, 1,416, 1,462, and 1,522 nm) was evidently correlated with the occurrence and development of diabetes. The number of hydrogen bonds was the smallest in the healthy group and the largest in the diabetes group. The suggested reason is that the water matrix of blood changes with the worsening of blood glucose metabolic dysfunction. The number of hydrogen bonds could be used as biomarkers for the early diagnosis of diabetes. The result show that it is effective and feasible to establish an accurate and rapid early diagnosis model of diabetes via NIRS combined with SVM and aquaphotomics.

show abstract

Model selection of SVMs using GA approach

Cited by 30 publications

References 11 publications

Deep learning classification of asteroids in g-type secular resonances

Deep learning classification of asteroids in g-type secular resonances

Two-Stage Genetic Algorhitm for Optimization Logistics Network for Groupage Delivery

Early Diagnosis of Type 2 Diabetes Based on Near-Infrared Spectroscopy Combined With Machine Learning and Aquaphotomics

Contact Info

Product

Resources

About