本文共 1768 字,大约阅读时间需要 5 分钟。
(Machine Learning Foundations)—Mathematical Foundations
,副教授 (Associate Professor),资讯工程学系 (Computer Science and Information Engineering)
Multiclass Classification 下的几种情况
Supervised Learning: every xn x n comes with corresponding yn y n
每个输入都知道对应的正确输出Unsupervised: Learning without yn y n
Semi-supervised: Learn with some yn y n
a very different but natural way of learning
惩罚错误的行为,奖励正确的行为
和机器的沟通方式
Batch Learning: 喂给机器一批一批的数据 (duck feeding)
监督学习、非监督学习Online Learning: 一个一个的来 (sequentially)
PLA、增强学习Active Learning: 主动的去学习 (ask questions)
总结
each dimension of X⊆Rd X ⊆ R d represents sophisticated physical meaning
具体的东西,可以计算,预先有人类智慧的加工
the easy ones for ML
更为抽象,包含很多细节 simple physical meaning
like image pixels, speech signal, etc.
often need human or machines to convert to concrete ones
需要机器自己去学到特征
no physical meaning
need feature conversion/extraction/construction