4: You trained four (4) deep neural networks for classifying music: D1 has 10 layers and a training errorof 0.03102, D2 has 5,000 layers and a training error of 0.00006, D3 has 500 layers and a training error of0.00023, and D4 has 3,000 layers and training error of 0.00014. Which network will be the most usefulfor classifying newly released music? Explain your