Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

L7-01 · Define the learning task

Separate quantum machine learning from causal identification.

Level 7 icon Level 7 · Doctoral & postdoctoral
Objective. Separate quantum machine learning from causal identification.
Before this lesson. L4-10, L6-10
Hands-on lab. Confounding and intervention

Learn

Quantum machine learning uses quantum methods in a learning procedure or learns from quantum data. Causal AI asks questions about interventions and mechanisms. Quantum causal inference studies causal questions about quantum systems. Their intersection is a research area, not one established universal algorithm.

Specify the input as classical records, quantum states, or access to a physical process. Specify the requested output as a label, expectation, channel estimate, graph, or causal effect. These choices determine the loss and the access assumptions. A classification score is not automatically a causal result.

A useful project statement names the estimand, data-generation mechanism, available interventions, and resource budget. An estimand is the quantity the study seeks to determine. A learning method can estimate it only if the model and experiment make that quantity identifiable.

Practice

Solution

Specify the candidate process family, available interventions, hidden-system assumptions, and identifiability conditions. Check that distinct claimed graphs have distinct allowed predictions. Prediction accuracy alone cannot establish those requirements.

Check your understanding

Which item defines a causal research target?

Show the answer

A. A precise intervention quantity and its assumptions

The target and its identification assumptions determine the causal question. A processor choice or prediction score alone does not supply them.

Sources

This lesson draws on Biamonte et al., 2017Costa & Shrapnel, 2016Huang et al., 2021.

Next: L7-02 · Variational circuits and parameter-shift gradients

References
  1. Biamonte, J., Wittek, P., Pancotti, N., Rebentrost, P., Wiebe, N., & Lloyd, S. (2017). Quantum Machine Learning. Nature, 549(7671), 195–202. 10.1038/nature23474
  2. Costa, F., & Shrapnel, S. (2016). Quantum causal modelling. New Journal of Physics, 18(6), 063032. 10.1088/1367-2630/18/6/063032
  3. Huang, H.-Y., Broughton, M., Mohseni, M., Babbush, R., Boixo, S., Neven, H., & McClean, J. R. (2021). Power of data in quantum machine learning. Nature Communications, 12(1). 10.1038/s41467-021-22539-9