机器学习定义：
A computer program is said to leran from experience E with respect to some task T and some performance measure P, if it's performance on T, as measured by P, improves with experience E.
例：
垃圾邮件分类，T：将邮件按是否为垃圾邮件分类；E：观察标记过的邮件是否为垃圾邮件；P：正确分类的邮件的数量。
监督式学习：回归、分类等。
无监督式学习：聚类（类如根据消费者信息的数据集自动地发现市场分类，并自动地把顾客划分到不同细分市场中）等。