2015
DOI: 10.1016/j.nima.2014.12.087
|View full text |Cite
|
Sign up to set email alerts
|

Neutron–gamma discrimination based on the support vector machine method

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

0
10
0

Year Published

2016
2016
2024
2024

Publication Types

Select...
8
1
1

Relationship

1
9

Authors

Journals

citations
Cited by 43 publications
(10 citation statements)
references
References 14 publications
0
10
0
Order By: Relevance
“…The data presented in Figure 10 are not linearly separable, therefore the linear combination presented is one of many possible solutions. Support vector machines are an approach to extend this work to include non-linear combinations by constructing a linear boundary in a transformed, higher-dimensional space [19]. Additionally, boosted decision trees are a popular means of classification that accurately map non-linear data relationships and easily adapt to additional parameters [20].…”
Section: Discussionmentioning
confidence: 99%
“…The data presented in Figure 10 are not linearly separable, therefore the linear combination presented is one of many possible solutions. Support vector machines are an approach to extend this work to include non-linear combinations by constructing a linear boundary in a transformed, higher-dimensional space [19]. Additionally, boosted decision trees are a popular means of classification that accurately map non-linear data relationships and easily adapt to additional parameters [20].…”
Section: Discussionmentioning
confidence: 99%
“…Integral ratio methods such as charge integral (CI) and partial charge-to-peak ratio (PCPR) methods are typical [20]. And machine learning methods include decision trees [21], support vector machines [22], K-nearest neighbors [23], artificial neural networks [24], and other approaches. The False Alarm Rate (FAR) values obtained using gamma sources can be utilized to evaluate the accuracy of the methods as mentioned above [20,25].…”
Section: Algorithms Of Pulse Shape Discriminationmentioning
confidence: 99%
“…Further work has been done to understand the performance of ANNs at specific energy ranges [10], to handle pile-up and high count rates [11], and to compare ANN performance in different detector types, like BC-501A and BC-537 [12]. Additionally, non-ANN forms of machine learning, such as Support Vector Machines, have been shown to be capable of particle-type discrimination in organic scintillators [13]. However, because the creation of supervised training and testing data sets relies on a priori knowledge of particle type, all such efforts have to contend with contamination from mis-classification, especially at low energies.…”
Section: B Machine Learningmentioning
confidence: 99%