Stochastic Gradient Descent with SVM for Imbalanced Data Classification
Journal Title: Scholars Journal of Physics, Mathematics and Statistics - Year 2016, Vol 3, Issue 4
Abstract
Stochastic Gradient Descent (SGD) is an attractive choice for SVM training. SGD leads to a result that the probability of choosing majority class is far greater than that of minority class for imbalanced classification problem. In order to deal with the large-scale imbalanced data classification problems, a method named stochastic gradient descent algorithm with SVM for imbalanced data classification is proposed. First, to deal with imbalanced data classification problems, we define the weight according to the size of positive and negative dataset. Then, a fast learning algorithm on large datasets called the weighted stochastic gradient descent algorithm with SVM is proposed, which helps to reduce the hyperplane offset to the minority class, thus solve the large-scale imbalanced data classification problems. Experimental results on real datasets show that the proposed method is effective.
Authors and Affiliations
Lu Shuxia, Zhu Chenxu, Zhou Mi
Discussion on a Kind of Sequence Limit
In this paper, we give four theorems and proved them. According to these four theorems, we deduce the solver method for the limit of a class of sequence { } by recursive relation .
A Note on the Identity Element in a Function Space
This note demonstrates that the identity element in appropriately defined function spaces is weakly compact, but not compact; and bounded, but not weakly compact.
Logistic Regression Modeling to Isolate Factors that Correlate with Usage of ITN as a Prophylactic to Malaria in Ghana
The study was conducted to isolate factors that correlate with ownership and usage of insecticide treated nets (ITNs) as a prophylactic to malaria in Asamankese, Ghana and explore the policy implications of the findings...
Ascertain Subclasses of Meromorphically Multivalent Functions with Negative Coefficient Associated with Linear Operator
In this paper, we introduce the subclasses and of meromorphic multivalent functions in the punctured unit disk by using a differential operator . We obtain coefficient estimates, distortion theorem, radius of convexi...
Comparison of criteria for the selection of discriminating variables: Application in Credit-Scoring
Banks want to reduce the credential risk by applying rules in order to classify the new loan seekers into “good customers” and “bad customers”. Searching past data is the best solution to build a statistics strategy to s...