Figure 4
The figure shows a neural network architecture that consists of three layers from left to right, labeled at the bottom as “Input Layer”, “Hidden Layer”, and “Output Layer”. The “Input Layer” contains four circles arranged vertically. There are three vertical dots between the first two circles and three vertical dots between the third and fourth circles. To the left of this layer, an arrow points from “D subscript t” to the top circle, and another arrow points from “D subscript t minus 12” to the bottom circle. The “Hidden Layer” is in the middle and grouped by a bracket below and consists of two sections. The first section on the left contains four circles with vertical three dots between the first and second circle, and vertical three dots between the third and fourth circle. This first set is connected to the “Input Layer” circles by multiple crossing arrows, where every circle in the “Input Layer” has an arrow pointing to every circle in the first section of the “Hidden Layer”. To the right of these circles, three dots (ellipses) lead to the second section of the “Hidden Layer”. This right section contains four circles arranged vertically, with “three vertical dots” appearing between the third and fourth circles. From each of the circles in this rightmost section of the “Hidden Layer”, a single arrow points toward a single circle present in the “Output Layer” on the right. To the right of the “Output Layer” circle, a final arrow points toward the label “C subscript t plus 6”.Structure of the DL model. Source: Authors