Research Article

Bidirectional Nonnegative Deep Model and Its Optimization in Learning

Figure 2

The unrolling illustration of BNDL.