Pearson Correlation-Based Feature Selection for Document Classification Using Balanced Training

. 2020 Nov 27;20(23):6793. doi: 10.3390/s20236793

Algorithm 1. Dataset balancing using a secondary dataset

Input:

D_{1}

Output:

D_{3}

Step 1:

X_{i} \leftarrow D_{1}

Step 2:

C_{i} \leftarrow C o u n t (X_{i}), w h e r e i = 1, \dots, n

Step 3:

T \leftarrow \max (C_{i})

Step 4:

D i f f_{i} \leftarrow {\begin{matrix} T - C_{i}, i f C_{i} < T \\ 0, i f C_{i} \geq T \end{matrix}

Step 5:

X_{j} \leftarrow F e t c h (D i f f_{i}, C_{i}, D_{2})

Step 6:

D_{3} = P o p u l a t e (D_{1}, X_{j})

return

D_{3}