Skip to content

Latest commit

 

History

History
53 lines (42 loc) · 2.46 KB

README.md

File metadata and controls

53 lines (42 loc) · 2.46 KB
Raw

QML Class Spring 2024 - Quantum Classifier Project

Project Overview

This project involves simulating a quantum classifier using a variational quantum circuit for binary classification problems. It is divided into three main parts, each contributing to the total project credits.

Part 1: Quantum Classifier for 3-input Parity Problem (6/10 Credits)

Requirements:

  • Objective: Build a quantum classifier for the binary classification parity problem with 3 inputs.
  • Data: Provided train and test datasets (classA_train.dat, classB_train.dat, classA_test.dat, classB_test.dat).
  • Data Preprocessing: Allowed to shift the data or perform simple preprocessing like subtracting 0.5 from all data.

Deliverables:

  1. Circuit Design: Design the quantum circuit including measurements (recommendation: do not exceed 4 qubits).
  2. Cost Function: Define the cost function using the expectation value of measurements at the end of the circuit.
  3. Optimization Method: Select and describe the optimization method used.
  4. Classification Outcomes: Report the accuracy and other relevant metrics for train and test data.
  5. Program Code: Provide the Python code used for the simulation.

Part 2: Extended Research (2/10 Credits)

Topics (choose one):

  • Compare different classical optimization methods and/or loss functions.
  • Justify the model using geometric representations like the Bloch sphere.
  • Compare parameters, epochs, and structure with a classical neural network solving the same problem.
  • Implement the parameter-shift rule for gradient evaluation.
  • Apply the classifier to another binary classification problem with 3 inputs and report the outcomes.
  • Any other interesting subject related to quantum machine learning (consultation with the instructor recommended).

Part 3: Extended Model for 5-input Parity Problem (2/10 Credits)

Requirements:

  • Objective: Extend the quantum classifier model to handle the parity problem with 5 inputs.
  • Data: Provided datasets (classA_train_N5.dat, classB_train_N5.dat, classA_test_N5.dat, classB_test_N5.dat).

Keywords

  • Quantum Machine Learning
  • Variational Quantum Circuit
  • Quantum Classifier
  • Binary Classification
  • Parity Problem
  • Quantum Optimization
  • Quantum Cost Function
  • Bloch Sphere
  • Parameter-shift Rule
  • Classical vs Quantum Models
  • Quantum Data Preprocessing
  • Quantum Simulation
  • Python Quantum Programming
  • Quantum Computing
  • Quantum Neural Networks