Algorithm3. K-nearest neighbor (KNN) algorithm.

Input: Label data set $D_{\mathrm{l}}={\{{\mathrm{x}}_{\mathrm{i}},y_{\mathrm{i}}\}}_{\mathrm{i}=1}^{\mathrm{l}}$    Pseudocode labeled unlabeled data set $D_{\mathrm{u}}={\left\{{\mathrm{x}}_{\mathrm{j}}\right\}}_{\mathrm{j}=1}^{\mathrm{u}}$    The parameter K in the KNN algorithm Process 1. for j = 1: u 2.   for i = 1: l 3.    Calculate the Euclidean distance ${\mathrm{S}}_{\mathrm{ji}}$ between $\left\{{\mathrm{x}}_{\mathrm{j}}\right\}$ and $\left\{{\mathrm{x}}_{\mathrm{i}}\right\}$ 4.   end 5.   Sort the labelled data set $D_{\mathrm{l}}$ in ascending order according to the distance ${\mathrm{S}}_{\mathrm{ji}}$ 6.   Select the first K data with tags, record their distance and tag information $\{s_1,y_1\},\cdots ,\left\{{\mathrm{s}}_{\mathrm{K}}{,\mathrm{y}}_{\mathrm{K}}\right\}$ 7.   for k = 1: K 8.    ${\mathrm{w}}_{\mathrm{k}}=\exp ({-\mathrm{s}}_{\mathrm{k}}^2/2)$ 9.   end    ${\mathrm{y}}_{\mathrm{i}}=\frac{{\displaystyle \sum _{\mathrm{k}=1}^{\mathrm{K}}{\mathrm{w}}_{\mathrm{k}}{\mathrm{y}}_{\mathrm{k}}}}{{\displaystyle \sum _{\mathrm{k}}^{\mathrm{K}}{\mathrm{w}}_{\mathrm{k}}}}$ 11. end 12.   Get pseudo-labelled data set ${\mathrm{D}}_{\mathrm{p}}{=\{\mathrm{x}}_{\mathrm{j}}{,\mathrm{y}}_{\mathrm{j}}{\}}_{\mathrm{j}=1}^{\mathrm{u}}$ Outputs: ${\mathrm{D}}_{\mathrm{knn}}{=\mathrm{D}}_{\mathrm{l}}\cup {\mathrm{D}}_{\mathrm{p}}$