[PR & ML 1] [Introduction] Informal Basic Concepts

最近还没更完OpenCV又开了新坑，谁教machine learning处在紧急又重要的地位呢。更新的内容总结自Pattern Recognition and Machine Learning by Christopher M. Bishop，英文书哪里都好，不过有时候表达一个意思要写好大一段啊，所以内容上只保留了精华部分。考虑应该做ML通用英文，所以没有翻译，文章中一些重要的“请读者证明”和练习用的Matlab代码也会一并更新。

Training phase (learning phase)

The process which determine the result function f(x) that takes the input x and generate an output prediction.

Generalization

The ability to categorize correctly new examples that differs from those  used for training. Generalization is a central goal in pattern recognition.

Feature extraction

The pre-process stage to transform the original input variables to some space of variables where, it is hoped the pattern recognition problem will be easier to solve or computation can be speeded up.

Pattern recognition problems

• Supervised learning problem: applications in which the training data comprises examples of the input vector along with their correponding target vectors
• Classification: the aim is to assign each input vector to one of a finite number od discrete categories
• Regression: the desired output consists of one or more continuous variables
• Unsupervised learning problem: the training data consist of a set of input vectors without any corresponding target values
• Clustering: to discover groups of similar examples within the data
• Density estimation: to determine the distribution of data within the input space
• Dimension reduction: to project the data from high-dimensional space down to low dimension
• Reinforce learning: the problrm of finding suitable actions to take in a given situation in order to maximize a reward. Here the learning algorithm is not given examples of optimal output, in contrast to supervised learning, but must discover them by a process of trails and errors