To utilize the structural information present in multidimensional features of an object, a tensor-based learning framework, termed as support tensor machines (STMs), was developed on the lines of support vector machines. In order to improve it further we have developed a least squares variant of STM, termed as proximal support tensor machine (PSTM), where the classifier is obtained by solving a system of linear equations rather than a quadratic programming problem at each iteration of PSTM algorithm as compared to STM algorithm. This in turn provides a significant reduction in the computation time, as well as comparable classification accuracy. The efficacy of the proposed method has been demonstrated in simulations over face detection and handwriting recognition datasets.
|Original language||English (US)|
|Number of pages||10|
|Journal||International Journal of Machine Learning and Cybernetics|
|State||Published - Dec 1 2013|
- Alternating projection
- Proximal support vector machines
- Support tensor machines