Introduction

1. Basic concepts of supervised learning

   • Sample, example, pattern
   • Features, predictors, independent variables
   • State of the nature, labels, pattern class, class, responses, dependent variables
     • \(\omega_1,\omega_2,\dots,\omega_c\) or \(y_1,y_2,\dots,y_c\) or \(z_1,z_2,\dots,z_c\)

   • Training data

     • \((x_1,\omega_1),(x_2,\omega_2),\dots,(x_n,\omega_n)\)

   • Model, statistical model, pattern class model, classifier

   • Test data
   • Training error & test error

2. classification(categorization, decision, making...): \(y\) is a categorical variable

3. regression: \(y\) is real-valued
4. Good representation: low intra-class variability and low inter-class similarity
5. cumulative distribution function(CDF): 累计分布函数
6. probability density function(PID): 概率密度函数

评论