Quantum Computing: A Data Science Approach
Project Description
An introductory, hands-on quantum computing project for data scientists, built in a single Colab notebook (Qiskit). It starts from intuition rather than math: why some problems scale exponentially and break classical computers, what a qubit and superposition actually are, and how interference lets quantum algorithms surface the right answer. From there, learners build real quantum circuits and implement two flagship algorithms — Grover’s search and QAOA for combinatorial optimization — before examining how real quantum hardware noise maps onto familiar data science problems. The framing throughout is that quantum concepts are the same ideas data scientists already know (search, optimization, parameterized models, noisy data) expressed in a different physical substrate. No physics background required; the notebook is designed to stretch you and runs entirely on a simulator.
Dataset
No external dataset required. All data is generated synthetically inside the notebook: search lists at increasing sizes (100 → 1,000,000) for the brute-force timing exercise, small toy instances for Grover’s oracle (e.g., a marked target state), and a small combinatorial optimization instance for QAOA. Qubit states and probability distributions are constructed directly in code.
Relevant Skills You May Apply
Python programming (loops, functions, timing), basic linear algebra and probability, algorithmic complexity / Big-O reasoning, comfort reading and running Jupyter/Colab notebooks.
Skills You May Gain
Quantum computing fundamentals (qubits, superposition, amplitudes, measurement, entanglement, the Bloch sphere), building and reading quantum circuits in Qiskit, implementing and reasoning about Grover’s algorithm (O(√N) search) and QAOA, and understanding quantum hardware noise (depolarizing, decoherence, readout error, crosstalk) through its data science analogues.
Total Time
Approximately 8–12 hours, self-paced. A few cells (QAOA optimization especially) take several minutes to run, which is expected.
Milestones
Milestone 1: Setup and Quantum Foundations — the case for quantum computing; implement and time a classical brute-force search to watch it break at scale (Step 4 TODO).
Milestone 2: Quantum Data Representation — a qubit as a probability distribution; bar-chart states and the Bloch sphere (no TODOs).
Milestone 3: Quantum Circuits — single-qubit gates, CNOT, multi-gate pipelines, product vs. entangled states, plus a build-your-own-circuit sandbox (no TODOs).
Milestone 4: Grover’s Algorithm — superposition, oracle, and diffuser; speedup analysis and a two-marked-state variation (written questions).
Milestone 5: QAOA — translating optimization objectives into a parameterized circuit; when approximate solutions are acceptable (written question).
Milestone 6: Quantum Hardware and Noise — the four noise types and how they map to data science problems (label noise, drift, miscalibration, correlated noise).
Deliverables
A completed notebook with the brute-force search implemented and timed (Milestone 1), all built circuits run and understood, and written answers for the Grover speedup/two-marked-state questions (M4) and the QAOA approximate-solution question (M5). Save a copy to your Personal Drive before starting.