Portrait of Mostafizur Rahaman

Mostafizur Rahaman

  • Researcher, KTH Royal Institute of Technology, Sweden
  • Research Scientist, IBM Quantum, on sabbatical
  • Visiting Scientist, RIKEN Center for Quantum Computing, Japan
Back to Teaching

Linear Algebra for Quantum Computing

Quantum computing is written in the language of linear algebra. These lectures cover the parts of it that are used throughout the subject: vectors for quantum states, matrices for quantum gates, and the ways matrices can be broken into simpler pieces.

Basic Quantum Programming (QT-03), AICTE Quantum Technology Programme, MNIT Jaipur, 2025 ยท Full playlist on YouTube

  1. 01

    Vector Spaces for Quantum Computing

    Day 7

    The state of a quantum system is a vector in a complex vector space. This lecture builds that language: vectors written in Dirac notation as kets and bras, the inner product that measures length and overlap, linear independence and bases. The computational basis of a qubit and the requirement that a state vector has unit length, so that measurement probabilities add up to one, connect the mathematics to quantum computing.

    • Complex vector spaces
    • Dirac notation
    • Kets and bras
    • Inner product
    • Norm and normalisation
    • Linear independence
    • Basis
    • Orthonormal basis
    • Computational basis

    Watch lecture

  2. 02

    Unitary Matrices and Quantum Gates

    Day 8, session 1

    Quantum gates are unitary matrices: matrices whose conjugate transpose is also their inverse. Because unitary matrices preserve the length of vectors, a gate always turns a valid quantum state into another valid one, and every gate can be undone. The lecture relates these properties to standard single-qubit gates such as the Pauli gates and the Hadamard gate.

    • Conjugate transpose
    • Unitary matrices
    • Hermitian matrices
    • Length preservation
    • Reversibility
    • Pauli gates
    • Hadamard gate
    • Phase gates

    Watch lecture

  3. 03

    Matrix Factorization Techniques

    Day 8, session 2

    A matrix factorisation writes a matrix as a product of simpler matrices, which exposes its structure. In the eigenvalue decomposition a matrix is described by its eigenvalues and eigenvectors. The spectral theorem guarantees that Hermitian and unitary matrices can be diagonalised this way using an orthonormal set of eigenvectors. The singular value decomposition extends the idea to any matrix, including rectangular ones.

    • Matrix factorisation
    • Eigenvalues and eigenvectors
    • Diagonalisation
    • Spectral theorem
    • Singular value decomposition

    Watch lecture