Support Vector Data Description Sampling for Fast Outlier Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The high computational time required to solve quadratic programming problems in Support Vector Data Description (SVDD) for large training datasets hinders efficient outlier identification.
Innovation Solution
A method involving randomly selecting multiple sets of observation vectors from the training dataset, computing optimal support vectors iteratively, and updating them until a stop condition is met, to define a support vector data description for outlier identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire training dataset is used to solve the quadratic programming problem, then the accuracy of outlier identification is improved, but the computational time increases significantly
Solution Approach 1:
The patent divides the training dataset into multiple smaller random samples. Instead of processing the entire dataset at once, the algorithm iteratively selects random samples (e.g., 10% of the data) and computes support vectors for each sample. This segmentation reduces the computational burden per iteration while maintaining the ability to capture outlier patterns through multiple iterations.
Solution Approach 2:
The patent uses partial action by processing only a subset (sample) of the training data in each iteration rather than the complete dataset. By repeatedly sampling and computing support vectors on these partial datasets, the algorithm achieves sufficient accuracy for outlier identification without the computational cost of processing all data simultaneously.
2Loss of time
If the number of support vectors is reduced by using sampling, then the computational time is reduced, but the data description accuracy may deteriorate
Solution Approach 1:
The patent implements continuity of useful action through iterative refinement. The algorithm repeatedly samples data, computes support vectors, and updates the SVDD model over multiple iterations. This continuous process allows the model to progressively improve its accuracy by incorporating information from different random samples, ensuring that the final data description is both accurate and computationally efficient.
Solution Approach 2:
The patent employs feedback mechanisms by comparing the SVDD models generated from different random samples and using this information to refine the final model. The algorithm evaluates the consistency and accuracy of support vectors across iterations and adjusts the model accordingly, ensuring that the reduced set of support vectors maintains adequate data description accuracy.
Data Source
AI summary
A computing device determines an SVDD to identify an outlier in a dataset. First and second sets of observation vectors of a predefined sample size are randomly selected from a training dataset. First and second optimal values are computed using the first and second observation vectors to define a first set of support vectors and a second set of support vectors. A third optimal value is computed using the first set of support vectors updated to include the second set of support vectors to define a third set of support vectors. Whether or not a stop condition is satisfied is determined by comparing a computed value to a stop criterion. When the stop condition is not satisfied, the first set of support vectors is defined as the third set of support vectors, and operations are repeated until the stop condition is satisfied. The third set of support vectors is output.