ACADEMIC PAPER SUMMARY

 

Paper summary:

 SUBVERTOR MACHINE LEARNING (SVM) is particularly suited to incremental learning for vast data classification since it has excellent power to summarize the data space. Based on considering whether a new data set can be incorporated into the history data, a heuristic incremental SVM learning algorithm is proposed. From partition difference of training data set, it collects more data points that contribute more to the FMALM hyperplane as support vectors. From partition difference of training data set, it collects more data points that contribute more to the FMALM hyperplane as support vectors. With the partition difference of training data, it collects more data points that contribute more to the FMALM hyperplane as support vectors. Experiments confirmed this algorithm is effective. The system is capable of handling vast data classification problems with a high degree of accuracy. There are promising results in this paper, and some additional studies will be done with a large amount and variety of data classification in the future.

Link for the paper: https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=1382061

Reference:

Zhong-Wei Li, Jian-Pei Zhang and Jing Yang, "A heuristic algorithm to incremental support vector machine learning," Proceedings of 2004 International Conference on Machine Learning and Cybernetics (IEEE Cat. No.04EX826), 2004, pp. 1764-1767 vol.3, doi: 10.1109/ICMLC.2004.1382061.