Optimizing Multidimensional Pooling for Variational Quantum Algorithms
Convolutional neural networks (CNNs) have proven to be a very efficient class of machine learning (ML) architectures for handling multidimensional data by maintaining data locality, especially in the field of computer vision. Data pooling, a major component of CNNs, plays a crucial role in extractin...
| Main Authors: | , , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-02-01
|
| Series: | Algorithms |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1999-4893/17/2/82 |
| _version_ | 1827344336681959424 |
|---|---|
| author | Mingyoung Jeng Alvir Nobel Vinayak Jha David Levy Dylan Kneidel Manu Chaudhary Ishraq Islam Evan Baumgartner Eade Vanderhoof Audrey Facer Manish Singh Abina Arshad Esam El-Araby |
| author_facet | Mingyoung Jeng Alvir Nobel Vinayak Jha David Levy Dylan Kneidel Manu Chaudhary Ishraq Islam Evan Baumgartner Eade Vanderhoof Audrey Facer Manish Singh Abina Arshad Esam El-Araby |
| author_sort | Mingyoung Jeng |
| collection | DOAJ |
| description | Convolutional neural networks (CNNs) have proven to be a very efficient class of machine learning (ML) architectures for handling multidimensional data by maintaining data locality, especially in the field of computer vision. Data pooling, a major component of CNNs, plays a crucial role in extracting important features of the input data and downsampling its dimensionality. Multidimensional pooling, however, is not efficiently implemented in existing ML algorithms. In particular, quantum machine learning (QML) algorithms have a tendency to ignore data locality for higher dimensions by representing/flattening multidimensional data as simple one-dimensional data. In this work, we propose using the quantum Haar transform (QHT) and quantum partial measurement for performing generalized pooling operations on multidimensional data. We present the corresponding decoherence-optimized quantum circuits for the proposed techniques along with their theoretical circuit depth analysis. Our experimental work was conducted using multidimensional data, ranging from 1-D audio data to 2-D image data to 3-D hyperspectral data, to demonstrate the scalability of the proposed methods. In our experiments, we utilized both noisy and noise-free quantum simulations on a state-of-the-art quantum simulator from IBM Quantum. We also show the efficiency of our proposed techniques for multidimensional data by reporting the fidelity of results. |
| first_indexed | 2024-03-07T22:45:40Z |
| format | Article |
| id | doaj.art-77c3de95f4d240959f6e166416220a69 |
| institution | Directory Open Access Journal |
| issn | 1999-4893 |
| language | English |
| last_indexed | 2024-03-07T22:45:40Z |
| publishDate | 2024-02-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Algorithms |
| spelling | doaj.art-77c3de95f4d240959f6e166416220a692024-02-23T15:04:31ZengMDPI AGAlgorithms1999-48932024-02-011728210.3390/a17020082Optimizing Multidimensional Pooling for Variational Quantum AlgorithmsMingyoung Jeng0Alvir Nobel1Vinayak Jha2David Levy3Dylan Kneidel4Manu Chaudhary5Ishraq Islam6Evan Baumgartner7Eade Vanderhoof8Audrey Facer9Manish Singh10Abina Arshad11Esam El-Araby12Department of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USADepartment of Electrical Engineering and Computer Science, University of Kansas, Lawrence, KS 66045, USAConvolutional neural networks (CNNs) have proven to be a very efficient class of machine learning (ML) architectures for handling multidimensional data by maintaining data locality, especially in the field of computer vision. Data pooling, a major component of CNNs, plays a crucial role in extracting important features of the input data and downsampling its dimensionality. Multidimensional pooling, however, is not efficiently implemented in existing ML algorithms. In particular, quantum machine learning (QML) algorithms have a tendency to ignore data locality for higher dimensions by representing/flattening multidimensional data as simple one-dimensional data. In this work, we propose using the quantum Haar transform (QHT) and quantum partial measurement for performing generalized pooling operations on multidimensional data. We present the corresponding decoherence-optimized quantum circuits for the proposed techniques along with their theoretical circuit depth analysis. Our experimental work was conducted using multidimensional data, ranging from 1-D audio data to 2-D image data to 3-D hyperspectral data, to demonstrate the scalability of the proposed methods. In our experiments, we utilized both noisy and noise-free quantum simulations on a state-of-the-art quantum simulator from IBM Quantum. We also show the efficiency of our proposed techniques for multidimensional data by reporting the fidelity of results.https://www.mdpi.com/1999-4893/17/2/82quantum computingconvolutional neural networksquantum machine learningpooling layers |
| spellingShingle | Mingyoung Jeng Alvir Nobel Vinayak Jha David Levy Dylan Kneidel Manu Chaudhary Ishraq Islam Evan Baumgartner Eade Vanderhoof Audrey Facer Manish Singh Abina Arshad Esam El-Araby Optimizing Multidimensional Pooling for Variational Quantum Algorithms Algorithms quantum computing convolutional neural networks quantum machine learning pooling layers |
| title | Optimizing Multidimensional Pooling for Variational Quantum Algorithms |
| title_full | Optimizing Multidimensional Pooling for Variational Quantum Algorithms |
| title_fullStr | Optimizing Multidimensional Pooling for Variational Quantum Algorithms |
| title_full_unstemmed | Optimizing Multidimensional Pooling for Variational Quantum Algorithms |
| title_short | Optimizing Multidimensional Pooling for Variational Quantum Algorithms |
| title_sort | optimizing multidimensional pooling for variational quantum algorithms |
| topic | quantum computing convolutional neural networks quantum machine learning pooling layers |
| url | https://www.mdpi.com/1999-4893/17/2/82 |
| work_keys_str_mv | AT mingyoungjeng optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT alvirnobel optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT vinayakjha optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT davidlevy optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT dylankneidel optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT manuchaudhary optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT ishraqislam optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT evanbaumgartner optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT eadevanderhoof optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT audreyfacer optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT manishsingh optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT abinaarshad optimizingmultidimensionalpoolingforvariationalquantumalgorithms AT esamelaraby optimizingmultidimensionalpoolingforvariationalquantumalgorithms |