ml-specialty video for a machine learning engineer noticed that a farm’s weeds are harmful to the actual healthy crops and decided to create a computer vision
A machine learning engineer noticed that a farm’s weeds are harmful to the actual healthy crops and decided to create a computer vision model to help distinguish weeds from crops. A convolutional neural network was used for this use case and the engineer trained the model for 3 consecutive days with data augmentation during training. At the end of the third day, the engineer noticed that the training accuracy is 93% while the validation and test accuracies were capped at 76%. Which techniques were to be applied in order to overcome this situation? (Select TWO.)