Difference
Transfer learning and Fine-tuning both replace the final layer to make predictions for a different task.
However, while transfer learning fixes all the layer previous to the final output layer, fine-tuning initializes the weights with the pre-trained weights but does not freeze them, making it more flexible.
