1: \begin{displaymath}

2: \label{Detail}

3: \setlength{\abovedisplayskip}{2pt}

4: \setlength{\belowdisplayskip}{2pt}

5: \setlength{\arrayrulewidth}{0.4mm}

6: \begin{tabular}{p{8.2cm}}

7: \hline

8: $\! \mathbf{Algorithm \ 1}$ Cross Domain Collaborative Learning \\

9: \hline

10: $\mathbf{Input}$: \\

11: \ \ \ Samples $\mathbf{X}_{l}^{s}$, $\mathbf{X}_{l}^{t}$ and $\mathbf{X}_{u}^{t}$. \\

12: \ \ \ Threshold for correlation coefficients: $\rho_T$=0.5.\\    

13: \ \ \ Initial training set: $\mathbf{TS} = \mathbf{X}_{l}^{t}$.\\

14: \ \ \ Free parameters for ERW: $\beta =710,\gamma=1e-5$.\\

15: \ \ \ Number of samples added into $\mathbf{TS}$: $p=10$.\\

16: 1: \textbf{Repeat}: \\

17: 2: \ \ Train a linear SVM classifier with probability esti-\\

18: \ \ \ \ \ mation using $\mathbf{TS}$.\\

19: 3: \ \ Classify $\mathbf{X}_{u}^{t}$ and estimate probability map $\widehat{\mathbf{P}}_1$.\\

20: 4: \ \ Update $\mathbf{TS}$ and $\mathbf{TCs}$ via RW-based pseudolabeling \\

21: \ \ \ \ \ using $\mathbf{TS}$ and $\widehat{\mathbf{P}}_1$.\\

22: 5: \ \ C-CCA using $\mathbf{X}_{l}^{s}$ and $\mathbf{TCs}$.\\

23: 6: \ \ Samples projected onto the subspace kept by $\rho_T$.\\

24: 7: \ \ Train a linear SVM classifier with probability esti- \\

25: \ \ \ \ \ mation using projected $\mathbf{X}_{l}^{s}$ and $\mathbf{X}_{l}^{t}$.\\

26: 8: \ \ Classify $\mathbf{X}_{u}^{t}$ and estimate probability map $\widehat{\mathbf{P}}_2$.\\

27: 9: \ \ Update $\mathbf{TS}$ via RW-based pseudolabeling using $\mathbf{TS}$ \\

28: \ \ \ \ \ and $\widehat{\mathbf{P}}_2$.\\

29: 10: \textbf{Until convergence}\\

30: 11: ERW-based classification using $\mathbf{TS}$ and $\widehat{\mathbf{P}}_2$.\\

31: 12: \textbf{Return} Classification map\\

32: \hline

33: \end{tabular}

34: \end{displaymath}

35: