Replication Data for: kernel l1-norm Principal Component Analysis for Denoising

This data set is created to evaluate the robustness of algorithm while de-noising nonlinear pattern data. The two-dimensional spiral data of the 3 cycle coordinates are (e^{0.14x}cos(x)+err,e^{0.14x}sin(x)+err), where err is sampled from a normal distribution (0,0.01) and x is a sequence of 476 numbers ranging from (-19,0) by step 0.04.

本数据集旨在评估算法在对非线性模式数据进行去噪时的鲁棒性。该数据集采用3周期坐标下的二维螺旋数据，其表达式为：(e^{0.14x}cos(x)+err, e^{0.14x}sin(x)+err)，其中err取自均值为0、标准差为0.01的正态分布，x为取值范围(-19, 0)、步长为0.04、共计476个数值的序列。