arXiv subjects

Aikaterini

Publications and source records attributed to Aikaterini.

3 recordsLinked to original sources

arXiv2022

The effect of the processing and measurement operators on the expressive power of quantum models

There is an increasing interest in Quantum Machine Learning (QML) models, how they work and for which applications they could be useful. There have been many different proposals on how classical data can be encoded and what circuit ans\"atze and measurement operators should be used to process the encoded data and measure the output state of an ansatz. The choice of the aforementioned operators plays a determinant role in the expressive power of the QML model. In this work we investigate how certain changes in the circuit structure change this expressivity. We introduce both numerical and analytical tools to explore the effect that these operators have in the overall performance of the QML model. These tools are based on previous work on the teacher-student scheme, the partial Fourier series and the averaged operator size. We focus our analysis on simple QML models with two and three qubits and observe that increasing the number of parameterized and entangling gates leads to a more expressive model for certain circuit structures. Also, on which qubit the measurement is performed affects the type of functions that QML models could learn. This work sketches the determinant role that the processing and measurement operators have on the expressive power of simple quantum circuits.

quant-ph↗

arXiv2021

Storage properties of a quantum perceptron

Driven by growing computational power and algorithmic developments, machine learning methods have become valuable tools for analyzing vast amounts of data. Simultaneously, the fast technological progress of quantum information processing suggests employing quantum hardware for machine learning purposes. Recent works discuss different architectures of quantum perceptrons, but the abilities of such quantum devices remain debated. Here, we investigate the storage capacity of a particular quantum perceptron architecture by using statistical mechanics techniques and connect our analysis to the theory of classical spin glasses. We focus on a specific quantum perceptron model and explore its storage properties in the limit of a large number of inputs. Finally, we comment on using statistical physics techniques for further studies of neural networks.

cond-mat.dis-nn↗

arXiv2021

Exploring Quantum Perceptron and Quantum Neural Network structures with a teacher-student scheme

Near-term quantum devices can be used to build quantum machine learning models, such as quantum kernel methods and quantum neural networks (QNN) to perform classification tasks. There have been many proposals how to use variational quantum circuits as quantum perceptrons or as QNNs. The aim of this work is to systematically compare different QNN architectures and to evaluate their relative expressive power with a teacher-student scheme. Specifically, the teacher model generates the datasets mapping random inputs to outputs which then have to be learned by the student models. This way, we avoid training on arbitrary data sets and allow to compare the learning capacity of different models directly via the loss, the prediction map, the accuracy and the relative entropy between the prediction maps. We focus particularly on a quantum perceptron model inspired by the recent work of Tacchino et. al. \cite{Tacchino1} and compare it to the data re-uploading scheme that was originally introduced by P\'erez-Salinas et. al. \cite{data_re-uploading}. We discuss alterations of the perceptron model and the formation of deep QNN to better understand the role of hidden units and non-linearities in these architectures.

quant-ph↗