There is a training dataset of $70,000$ images of different articles of clothing. Each image should be assigned to one of ten possible classes, for example: $$shirt$,$sneaker$,$boot$.$ The training dataset has exactly $7,000$ images for each class.

You want use this dataset to train a neural network to classify an image of an article of clothing. Before the neural network is trained it obviously cannot classify well. The untrained neural network randomly guesses a class for any image it is shown, and each of the ten classes is equally likely to be the guess.

The neural network is provided one image from the training dataset, and the neural network guesses the class for that image. What is the probability that the class guessed by the neural network is the actual class of the image?

Express the probability as a percent, i$.$e$.$ a number between $0$ and $100,$ where $100$ means a $100\%$ probability that the class guessed by the network is the correct class, and $0$ means a $0\%$ probability that the class guessed by the network is the correct class.