
这里先来复习一下伟大的贝叶斯(Bayesion)公式:
贝叶斯公式说明了条件概率里面原因和结果概率之间的联系。贝叶斯分类器先设定一组条件的先验概率,然后用一组数据对分类器进行训练,得到更加精确的后验概率,最后就可以用分类器来进行分类和预测了。
今天,极天公司来实验室做交流,介绍了Autonomy公司,一家企业搜索提供商。极天公司是Autonomy的OEM商,为企业提供企业搜索解决方案(主要技术核心应该是在Autonomy的搜索引擎之上做了基于lucene的中文划词插件)。
期间,对Autonomy公司产生了兴趣,google之,发现了一个传奇的CEO, Lynch博士。他从大学期间开始创业,先后成功创办了多家高科技公司,其中3家以上的公司上市,最为成功的就是号称要Beat Google Up的Autonomy。从Lynch先生的生平我们可以推断,Autonomy公司的搜索引擎应该是基于Bayesion分类器的(Lynch先生从研究生开始就在搞Bayesion模式识别),他创办的几家公司都和神经网络、模式识别有关。
-
Michael Richard Lynch - Wikipedia, the free encyclopedia
- again based on his Bayesian research
- Shortly after the dot com crash he upset some in the city and won praise from others in an infamous incident.
-
-
Dr Mike Lynch is the chief executive officer of Autonomy Corporation plc,
- Dr Mike Lynch is the chief executive officer of Autonomy Corporation plc,
-
-
- Mike Lynch
- His first company was Cambridge Neurodynamics, which he set up in 1991.
- getting computers to recognise things - and solved a big problem in the area of neural networks
-
Posted from Diigo. The rest of my favorite links are here.
没有评论:
发表评论