基本信息
浏览量:87
职业迁徙
个人简介
Craig Saunders is currently a lecturer in the Image, Speech and Intelligent Systems Research Group, which is part of the School of Electronics and Computer Science at the University of Southampton, which he joined in January 2004. Previously he was in the Computer Science department at Royal Holloway, University of London where he was a lecturer from September 2000 to December 2003. Craig Saunders also obtained his undergraduate degree and PhD from Royal Holloway, completing a degree Computer Science and Management studies in 1997 and a PhD in Machine Learning in 2000.
Craig's main interest is in the development and application of machine learning methods. His PhD focussed on using approximations to universal randomness tests to provide rigourous confidence values for machine learning methods. This was based largely on the typicalness framework of Vovk, where Martin-Loef and Levin tests are used as typical universal approximations. These were used in conjunction with kernel methods, particularly SVMs and Ridge Regression, to provide confidence measures rather than standard bare predictions. During his PhD he was also involved in the Royal Holloway SVM software -- largely written by Mark Stitson and Jason Weston, but contributed to by many -- which was one of the first widely-used SVM implementations. This has now been superseded by many good SVM and more general toolboxes.
More recently his research work has centered around developing kernels for discrete data. This has included work on text analysis using variations of the string kernel, and in studying the connections between current structure kernels and fisher-style kernels which use underlying probabilistic models. He is currently working on developing methods for bioinformatics and chemoinformatics, developing probabilistic models and kernels for protein sequences and molecular structures. He is also interested in the problem of feature selection via subspace methods such as kernel PCA and kernel Partial Least Squares. These methods tend to go hand-in-hand with stucture kernels, as the latter usually results in very high dimensional (and noisy) feature spaces. He is currently investigating a general framework for these methods, and is looking at sparse greedy approximations to them.
He also has a strong interest in strategic game playing, and games in general. His interests particularly lie with the oriental game of Go (which he is still struggling to improve at) and the huge research challenges that the game provides for AI and Machine Learning researchers.
研究兴趣
论文共 75 篇作者统计合作学者相似作者
按年份排序按引用量排序主题筛选期刊级别筛选合作者筛选合作机构筛选
时间
引用量
主题
期刊级别
合作者
合作机构
mag(2015)
引用24浏览0引用
24
0
mag(2013)
引用33浏览0引用
33
0
mag(2013)
引用235浏览0引用
235
0
加载更多
作者统计
合作学者
合作机构
D-Core
- 合作者
- 学生
- 导师
数据免责声明
页面数据均来自互联网公开来源、合作出版商和通过AI技术自动分析结果,我们不对页面数据的有效性、准确性、正确性、可靠性、完整性和及时性做出任何承诺和保证。若有疑问,可以通过电子邮件方式联系我们:report@aminer.cn