A variational quantum perceptron with Grover’s algorithm for efficient classification

Abstract

Abstract This study introduces the Quantum Variational Perceptron with Grover’s algorithm (QVP-G), an innovative Quantum machine Learning (QML) model significantly enhancing binary classification tasks’ capabilities. The study goes beyond theoretical constructs, employing empirical evaluations of QVP-G on the well-established Iris and more complex breast cancer datasets. This dual-dataset approach highlights the model’s adaptability and effectiveness in varied contexts. Our experimental findings reveal the exceptional accuracy of QVP-G, achieving a remarkable 99% success rate. This superior performance is owed mainly to integrating Grover’s algorithm, a cornerstone in Quantum Computing (QC), which substantially increases the probability of accurately determining the correct classification hyperplane. Beyond academic interest, our research positions QVP-G as a powerful tool with substantial practical applications. Its utility ranges from enhancing fraud detection and improving spam filtering to expediting drug discovery and contributing to advances in physics research. This study marks a significant advancement in QML, reinforcing the bridge between QC and machine learning.