Quantum Kernels for Real-World Predictions Based on Electronic Health Records

Abstract

Research on near-term quantum machine learning has explored how classical machine learning algorithms endowed with access to quantum kernels (similarity measures) can outperform their purely classical counterparts. Although theoretical work has shown provable advantage on synthetic data sets, no work done to date has studied empirically whether quantum advantage is attainable and with what data. In this paper, we report the first systematic investigation of empirical quantum advantage (EQA) in healthcare and life sciences and propose an end-to-end framework to study EQA. We selected electronic health records (EHRs) data subsets and created a configuration space of 5-20 features and 200-300 training samples. For each configuration coordinate, we trained classical support vector machine (SVM) models based on radial basis function (RBF) kernels and quantum models with custom kernels using an IBM quantum computer, making this one of the largest quantum machine learning experiments to date. We empirically identified regimes where quantum kernels could provide advantage and introduced a terrain ruggedness index, a metric to help quantitatively estimate how the accuracy of a given model will perform. The generalizable framework introduced here represents a key step towards a priori identification of data sets where quantum advantage could exist.INDEX TERMS artificial intelligence, digital health, electronic health records, empirical quantum advantage, machine learning, quantum kernels, real-world data, small data sets, support vector machines

Quantum computing algorithms: getting closer to critical problems in computational biology

“…Of course, the final interpretation or evaluation of the biological meaning of a difference/variation observed remains the operator role, but in this condition, the difference is detected by the machine autonomously, automatically and objectively. As concerns quantum approaches, they have been proposed to improve the classification process [ 58 , 59 ], either by development of quantum neural networks [ 60–62 ] or, more recently, by development of a quantum support vector machine [ 63 ]. Several approaches have been proposed to train a network using QC technology in a more accurate, robust and quick way.…”

Robust Machine Learning Systems: Challenges,Current Trends, Perspectives, and the Road Ahead

“…Constructing QKs for SVMs has been under investigation in recent works. The performance of the resulting quantum SVMs for classification problems with a small number of training points exceeds that of optimized classical models with conventional kernels [85] [86]. Moreover, QKs allow us to interpret the learning process with the help of quantum information tools [87].…”

Linear, Bidirectional and Circular Controlled Quantum Teleportation and Quantum State Sharing using a Seven Qubit Genuinely Entangled State

“…Quantum computing can have absolute benchmarks like the scaling of run-time while machine learning have more constructivist benchmarks between disparate models. Quantum advantage can be spoken of in terms of solvability, expressivity of the class of the model, size of training sample needed, generalizability and how the optimization landscape is structured (162)(163)(164)(165)(166)(167)(168)(169)(170)(171). Due to the need to be able to capture the system dynamics in terms of quantum systems and circuitry, we can only gauge performance of machine learning algorithms in specific selected contexts and examples, with this selectivity preventing any generalizability of performance-based advantage, if any.…”