Second Order Online Active Learning based Malicious Web Classification

Iterative learning algorithms that approximate the solution of support vector machines (SVMs) have two potential advantages. First, they allow online and active learning. Second, for large data sets, computing the exact SVM solution may be too time-consuming, and an efficient approximation can be preferable. The powerful LASVM iteratively approaches the exact SVM solution using sequential minimal optimization (SMO). It allows efficient online and active learning. Here, this algorithm is considerably improved in speed and accuracy by replacing the working set selection in the SMO steps. A second-order working set selection strategy, which greedily aims at maximizing the progress in each single step, is incorporated.

Download Full-text

SOAL: Second-Order Online Active Learning

2016 IEEE 16th International Conference on Data Mining (ICDM) ◽

10.1109/icdm.2016.0115 ◽

2016 ◽

Cited By ~ 5

Author(s):

Shuji Hao ◽

Peilin Zhao ◽

Jing Lu ◽

Steven C. H. Hoi ◽

Chunyan Miao ◽

...

Keyword(s):

Active Learning ◽

Second Order

Download Full-text

Second-Order Online Active Learning and Its Applications

IEEE Transactions on Knowledge and Data Engineering ◽

10.1109/tkde.2017.2778097 ◽

2018 ◽

Vol 30 (7) ◽

pp. 1338-1351 ◽

Cited By ~ 4

Author(s):

Shuji Hao ◽

Jing Lu ◽

Peilin Zhao ◽

Chi Zhang ◽

Steven C.H. Hoi ◽

...

Keyword(s):

Active Learning ◽

Second Order

Download Full-text

Disappearance Voltage Determinations Using a Convergent Beam

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100064979 ◽

1971 ◽

Vol 29 ◽

pp. 184-185

Author(s):

W. L. Bell

Keyword(s):

Second Order ◽

Objective Method ◽

Uniform Thickness ◽

Rocking Curve ◽

Crystal Perfection ◽

Kikuchi Line ◽

Central Maximum ◽

Diffraction Patterns ◽

Convergent Beam ◽

Beam Technique

Disappearance voltages for second order reflections can be determined experimentally in a variety of ways. The more subjective methods, such as Kikuchi line disappearance and bend contour imaging, involve comparing a series of diffraction patterns or micrographs taken at intervals throughout the disappearance range and selecting that voltage which gives the strongest disappearance effect. The estimated accuracies of these methods are both to within 10 kV, or about 2-4%, of the true disappearance voltage, which is quite sufficient for using these voltages in further calculations. However, it is the necessity of determining this information by comparisons of exposed plates rather than while operating the microscope that detracts from the immediate usefulness of these methods if there is reason to perform experiments at an unknown disappearance voltage.The convergent beam technique for determining the disappearance voltage has been found to be a highly objective method when it is applicable, i.e. when reasonable crystal perfection exists and an area of uniform thickness can be found. The criterion for determining this voltage is that the central maximum disappear from the rocking curve for the second order spot.

Download Full-text