Condensed Matter Theory Seminar | October 07, 14:00

Tensor network algorithms for 1D spin chains: DMRG and entanglement entropy minimization

Leon Schoonderwoerd

Tensor networks provide a novel and powerful way to think about and perform calculations with many-body systems. This talk treats tensor networks in the form of Matrix Product states (MPS) and Matrix Product Operators (MPO), which are used for calculations on (1D) spin chains. Firstly, Density Matrix Renormalization Group (DMRG) algorithms, concerned with finding the ground state and its energy. Secondly, an approach for the minimization of entanglement entropy using unitary basis transformations is given. The idea for this latter part is based on [1], but instead of a parameter-based approach, an environment network-based algorithm is used. The combination of the entropy minimization with DMRG is discussed. Results obtained using the minimization algorithm are shown, and the method's limitations are discussed. [1] C. Krumnow, L. Veis, Ö. Legeza & J. Eisert, Fermionic orbital optimisation in tensor network states, arXiv 1504.00042v3


University of Amsterdam
Seminar Room 0.03, ETP
Contact: Simon Trebst