Detail — Mlp

Metrics

• Accuracy: 96.23%
• Macro F1: 96.08%
• Weighted F1: 96.23%
• Best Val Acc: 97.64%

Config (JSON)

{
  "tag": "narrow_many_4layers",
  "layers": [
    384,
    256,
    192,
    128
  ],
  "dropouts": [
    0.1,
    0.15,
    0.2,
    0.2
  ],
  "input_dropout": 0.05,
  "activation": "relu",
  "optimizer": "adam",
  "lr": 0.0008,
  "epochs": 160,
  "batch": 256,
  "use_smote": true,
  "test_size": 0.15,
  "val_size": 0.15,
  "early_stop_patience": 20,
  "reduce_lr": true,
  "rlr_factor": 0.5,
  "rlr_patience": 8
}

Source MD

md/mlp_narrow_many_4layers.md