Algorithm 1: Our Modified Residual-Dense CNN Network Architecture
Step 1:	Input: Training dataset, X; Corresponding labels, Y; Number of residual blocks, num_res_blocks Number of dense blocks, num_dense_blocks; Number of layers within each residual block, num_res_layers; Number of layers within each dense block, num_dense_layers
Step 2:	Output: Trained Residual-Dense Network model
Step 3:	[Initialization]: a. Define the input layer: Input layer: $X_{Input}$, shape determined by dataset b. Add an initial convolutional layer: Initial Convolutional Layer: $X_{conv}=Conv1D\left(X_{Input}, filters=63, kernel=3\right)$ c. Add residual-dense blocks: Initialize $X_{current}=X_{conv}$
Step 4:	[Define Inner Residual Block]: Repeat For k = 1 to num_res_blocks do: a. Initialize a new Block: Block input: $X_{block\_input}=X_{current}$ b. Repeat For I = 1 to num_res_layers do Convolutional layer: Convolutional Layer:   $X_{Conv\_i}=Conv1D\left(bloc{k}_{Input}, filters=63, kernel=3\right)$ Batch normalization:  $Bn\_i=BatchNormalization\left(X_{Conv\_i}\right)$ Activation:  $ac{t}_i=ReLU\left(B{n}_i\right)$ c. Residual connection: Residual output: $X_{residual}=X_{act\_num\_res\_layers}+X_{Current}$ d. Update $X_{current}=X_{residual}$
Step 5:	[Define the Dense Block]: Repeat For k =1 to num_dense_blocks do: a. Initialize a new Block: Block input: $X_{block\_input}=X_{current}$ b. Repeat For I= 1 to num_dense_layers do Convolutional layer: Convolutional Layer: $X_{Conv\_i}=Conv1D\left(bloc{k}_{Input}, filters=63, kernel=3\right)$ Batch normalization:  $Bn\_i=BatchNormalization\left(X_{Conv\_i}\right)$ Activation:  $ac{t}_i=ReLU\left(B{n}_i\right)$ c. Concatenate with block input: output: $X_{concat}=\mathrm{Concatenate} \left(X_{bloc{k}_{input}},X_{num\_dense\_layers}\right)$ d. Update $X_{current}=X_{concat}$
Step 6:	Global Average Pooling and Fully Connected Layers: Global average pooling: $avg\_pool=GlobalAvgPooling(X_{current}$) Flatten layer: $flt\_layer=Flatten(avg\_poo{l}_{ })$ Fully-connected layers: $dense=Dense(flt\_layer{,}_{ }units=64, activation=ReLU$)
Step 7:	[Return Trained Residual-Dense Network model]