Model Performance

Model comparison

Two convolutional architectures evaluated on the validation set. Higher is better on every metric.

Side-by-side metrics

Best: Deeper CNN

Metric	Simple CNN	Deeper CNN
Accuracy	0.8213	0.8493
F1 (weighted)	0.8171	0.8416
Precision (weighted)	0.8252	0.854
Recall (weighted)	0.8213	0.8493

Training curves

Accuracy and loss per epoch

Training Curves

Performance bar chart

Weighted metrics

Model Comparison

Confusion matrices

Validation set

Simple CNN

Simple CNN confusion matrix

Deeper CNN

Deeper CNN confusion matrix