Quantum Machine Learning (QML) has emerged as a promising paradigm that integrates the computational advantages of quantum computing with the versatility of classical machine learning techniques. This study proposes a hybrid quantum classical framework for numerical regression tasks, leveraging the capabilities of variational quantum circuits to capture complex, high-dimensional relationships within numerical datasets. The methodology encodes numerical features into quantum states using amplitude encoding and applies a parameterized quantum circuit trained via a classical optimizer. Experimental evaluation is conducted using IBM Qiskit's quantum simulator on benchmark regression datasets, comparing model performance against classical linear regression and neural network baselines. Results demonstrate that the proposed hybrid approach achieves competitive prediction accuracy while reducing model complexity, particularly for small to mediumsized datasets. Furthermore, the study discusses computational resource requirements, noise resilience, and scalability challenges, offering insights into the practical adoption of QML in predictive analytics.
Quantum Machine Learning, Numerical Regression, Variational Quantum Circuits, Hybrid QuantumClassical Computing, Qiskit, Predictive Modeling, Amplitude Encoding, Quantum Algorithms, Regression Analysis, Quantum Data Encoding.