Experiment 5
Binary classification with validation set implementation.
Parameters
basic setting parameters
clip_length: 5.0 # [sec]
preprocessing parameters
sample_rate: 44100
hop_length: 512
n_fft: 1024
n_mels: 64
Training parameters
number of audio samples: 220500
learning rate: 0.001
batch size: 130
number of epochs: 30
number of samples: 1040
balanced dataset: False
random clip cut: False
scheduler: True
Classes
Labels
grime
jazz
Class distribution
Dataset | Class | Class ID | Samples |
---|---|---|---|
Train | grime jazz | [0, 1] | 1040 |
Validation | grime jazz | [0, 1] | 160 |
Test | grime jazz | [0, 1] | 400 |
Metrics
ID | Epochs | Batch | Classes | LR | samples | Loss | Accuracy |
---|---|---|---|---|---|---|---|
30 | 130 | 2 | 0.001 | 1040 | 0.15 | 0.94 |
Loss
![](/_next/static/media/loss-experiment-Feb27_02-30-45.f13d2686.png)
Accuracy
![](/_next/static/media/acc-experiment-Feb27_02-30-45.dcf3a0ef.png)
Class Accuracy
![](/_next/static/media/class-acc-experiment-Feb27_02-30-45.84abe559.png)
Pr Curves
![](/_next/static/media/batch-experiment-Feb27_02-30-45.d63dc486.png)
Pr Curves
![](/_next/static/media/pr-curvesexperiment-Feb27_02-30-45.89912257.png)
Confusion Matrix
![](/_next/static/media/cf-curvesexperiment-Feb27_02-30-45.ec34cb2f.png)