I am working on a project related to Boosting of Quantum Classifiers of PAC format. And I am a little confused about how do we go about implementing a PAC classifier. The basic idea is that we have to use multiple copies(say Q) of the state $$ \sum_{x \in \{ 0,1 \}^n } \sqrt{D(x_i)} |x_i, y_i = c(x_i)> $$

where in order to train our classifier. On further exploring the paper pointed out that we need to measure out the above state and then use the samples in order to train the classifier.

So assuming that I have access to such a state, would measuring out the state and then using Q states out of all the obtained states in order to train a classifier such as QNN or QSVM, accomplish the said task.

Furthermore lets say I have the $ D(x_i)$ and $(x_i, y_i)$ is there even any need to create the said state, or just directly using these to train our model enough ??



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.