Application of quantum machine learning using quantum kernel algorithms on multiclass neuron M-type classification.

Area of Science:

• Neuroscience
• Quantum Computing
• Machine Learning

Background:

• Functional characterization of neuronal types is a critical challenge.
• Quantum machine learning (QML) offers potential for novel computational approaches.
• Previous QML studies showed promise on artificial datasets and small binary problems.

Purpose of the Study:

• To investigate the performance of quantum systems for multiclass neuron morphology classification using real-world data.
• To evaluate the efficacy of quantum kernel methods in this domain.
• To explore the impact of feature engineering on classification accuracy.

Main Methods:

• Application of quantum kernel methods for automatic multiclass neuron classification.
• Utilizing real-world neuron morphology datasets.
• Analysis of feature engineering's influence on classification performance.

Main Results:

• Quantum kernel methods achieved performance comparable to classical methods for neuron morphology classification.
• Certain configurations demonstrated advantages for quantum approaches.
• Feature engineering significantly influenced classification accuracy.

Conclusions:

• Quantum kernel methods are a viable approach for multiclass neuron classification.
• Further research is warranted to fully leverage quantum advantage with real-world neuroscience data.
• This study pioneers the use of quantum systems for classifying neuron morphologies.