Simulator-only study of when amplitude-encoded variational quantum classifiers fail, with validated fixes. Real amplitude encoding obeys rho(x) = rho(-x). When preprocessing (typically z-scoring) drives class means antipodal, encoded classes collapse and a loss barrier appears. This work provides a one-pass pre-training risk score (class-mean cosine, with class-mean trace distance T and a linear probe) plus three A/B-tested symmetry-breaking remedies. Headline: class-mean cosine predicts classical-minus-QNN balanced-accuracy gap across 15 public binary datasets at Spearman rho = -0.72 (dataset-permutation p = 0.0034), with 10 datasets held out from calibration. Fixed binary cutoffs do not transfer as universal constants; the transferable instrument is the continuous score. Remedies recover selectively on flagged datasets (breast cancer 0.784 → 0.961, d_z up to 6.5). Synthetic sweep isolates mechanism (Spearman 1.0, classical at ceiling). All exact statevector, verified to ~1e-7. Contents of this upload (A1_arXiv_bundle.zip):- A1_manuscript.md, appendix theory, related work, references- figures/: 5 PNGs (300 dpi)- tables/: diagnostic tables, remedy A/B CSVs, correlation JSONs for Tables 1,2,5-9- code/: src/, run_experiment.py, key reproduce scripts, requirements.txt Reproduce headline: python scripts/balanced_accuracy_validation.pyFull test suite in source repo: python -m pytest (191 passed). Author: Muhammad Huzaifa Farooq, Independent Researcher. Contact: huzaifa.farook@gmail.com. No specific funding. AI assistance (Claude Code) used for drafting/coding, all results verified by author.
Simulator-only study of when amplitude-encoded variational quantum classifiers fail, with validated fixes. Real amplitude encoding obeys rho(x) = rho(-x). When preprocessing (typically z-scoring) drives class means antipodal, encoded classes collapse and a loss barrier appears. This work provides a one-pass pre-training risk score (class-mean cosine, with class-mean trace distance T and a linear probe) plus three A/B-tested symmetry-breaking remedies. Headline: class-mean cosine predicts classical-minus-QNN balanced-accuracy gap across 15 public binary datasets at Spearman rho = -0.72 (dataset-permutation p = 0.0034), with 10 datasets held out from calibration. Fixed binary cutoffs do not transfer as universal constants; the transferable instrument is the continuous score. Remedies recover selectively on flagged datasets (breast cancer 0.784 → 0.961, d_z up to 6.5). Synthetic sweep isolates mechanism (Spearman 1.0, classical at ceiling). All exact statevector, verified to ~1e-7. Contents of this upload (A1_arXiv_bundle.zip):- A1_manuscript.md, appendix theory, related work, references- figures/: 5 PNGs (300 dpi)- tables/: diagnostic tables, remedy A/B CSVs, correlation JSONs for Tables 1,2,5-9- code/: src/, run_experiment.py, key reproduce scripts, requirements.txt Reproduce headline: python scripts/balanced_accuracy_validation.pyFull test suite in source repo: python -m pytest (191 passed). Author: Muhammad Huzaifa Farooq, Independent Researcher. Contact: huzaifa.farook@gmail.com. No specific funding. AI assistance (Claude Code) used for drafting/coding, all results verified by author.
Variational quantum classifiers (VQCs) are often compared with classical models under unequal data partitions, tuning budgets, or statistical treatment. We benchmark a 4-qubit VQC against logistic regression, an RBF-SVM, random forest, a quantum-kernel SVM (QSVM), and a majority-class control on Breast Cancer Wisconsin (n=569), Coimbra (n=116), and Golub leukemia (n=72), using five stratified folds across five seeds (25 paired blocks per dataset), identical training-fold-fitted preprocessing, and Nadeau–Bengio corrected paired tests with Holm correction. A first arm with a frozen configuration and the common [0, π] angle encoding left the VQC at 0.563, 0.515, and 0.576 accuracy (majority 0.617, 0.552, 0.652). In a second arm, in which every model was tuned by the same inner cross-validation and the encoding scale was a tuned hyperparameter, VQC accuracy rose to 0.864, 0.616, and 0.747. On Wisconsin the VQC stayed significantly below every matched classical model (differences 0.079–0.097, Holm p≤0.006) but above the majority class; on Golub it was significantly below only the SVM (Holm p=0.016); on Coimbra no comparison was significant. A QSVM with the same feature map exceeded the VQC numerically on every dataset, and three optimizers reached the same training loss in pilot diagnostics, pointing to the encoding and the variational head rather than the optimizer. The Wisconsin arms also differ in sample size (120 versus 569). Results apply to this low-qubit, PCA-compressed, exactly simulated setting and do not indicate a quantum advantage.
Nowadays, quantum-enhanced methods have been widely studied to solve machine learning related problems. This article presents the application of a Variational Quantum Classifier (VQC) for binary classification. We utilized three datasets: a synthetic dataset with randomly generated values between 0 and 1, the publicly available University of California Intelligence Machine learning (UCI) sonar dataset consisting of mining data, and a proprietary diabetes dataset related to diabetes with acute diseases and diabetes without acute disease. To deal with the limitation of noisy intermediate-scale quantum systems (NISQ), we used a pre-processing method to enhance the prediction rate when applying the VQC method. The process includes feature selection and state preparation. Quantum state preparation is critical for obtaining a functioning pipeline in a quantum machine learning (QML) model. Amplitude encoding is a state preparation approach that enhances the performance of data encoding and the learning of quantum models. As a result, our proposed methods achieved accuracies of 75%, 71.4%, and 68.73% by using VQC model and in contrast, the amplitude encoding-based VQC achieved 98.40%, 67.3%, and 74.50% accuracies on the synthetic, sonar, and diabetes dataset, respectively.
Comparisons of quantum data-encoding strategies are rarely accompanied by a registered estimand, an explicit treatment of where simulated noise is applied, or a classical reference point. This study reports a registered simulation study of the interaction between data encoding (angle versus ZZ feature map) and circuit noise (noiseless versus single-Pauli errors with probability p = 0.02 after every CNOT, plus readout flips) for an 8-qubit variational quantum classifier trained with SPSA. Three binary datasets, ten confirmatory seeds (240 runs) and circuit size as a secondary factor were used. Hyperparameters were tuned symmetrically on separate seeds and frozen before the confirmatory runs. The registered primary estimand, the Encoding x Noise interaction on validation balanced accuracy, was -0.154 (95% seed-cluster bootstrap CI [-0.181, -0.124]), which supports the registered directional hypothesis. ZZ-encoded models trained under noise met the pre-declared convergence criterion in 1 of 60 runs, against 48 of 60 for angle encoding, so training failure and signal degradation are not separated by the registered design. An exploratory follow-up, in which ZZ models were trained without noise and evaluated under noise, showed accuracy falling to 0.55-0.63 at p = 0.02, which points to the number of noisy two-qubit gates (56 additional CNOTs) as the main driver. The registered transfer hypothesis was not supported: robust selection did not differ from naive selection (+0.003, CI [-0.012, +0.016]). A class-balanced logistic regression exceeded every quantum configuration on validation accuracy. No quantum advantage is claimed.
Variational quantum classifiers (VQCs) are often reported to perform poorly on biomedical tabular data, but such conclusions depend on the encoding circuit, the training budget, and the comparison protocol. Methods. We compared 4-qubit VQCs, a quantum-kernel SVM (QSVM), a majority-class control, and tuned classical models (on four PCA features and on all features) on Wisconsin breast cancer (n = 569), Coimbra biomarkers (n = 116), Golub leukemia gene expression (n = 72, 7,129 genes), and SEER breast-cancer registry data (n = 4,024; 15% deaths). All models were evaluated on identical blocks of 5-fold × 5-seed stratified cross-validation with preprocessing fitted on training folds, nested tuning, and Nadeau–Bengio-corrected inference with Holm correction. Trained classifiers were also evaluated on three IBM Quantum devices. Results. On the three small datasets, a VQC with the conventional two-repetition ZZ feature map was close to chance (accuracy 0.688, 0.594, 0.509), whereas a product (Z) feature map reached 0.900, 0.797, and 0.658; at matched repetition counts the Z-minus-ZZ gap was +0.12 to +0.21 and, after Holm correction, significant on Wisconsin only. More evaluations (up to 1,500) and six initializations did not close the ZZ deficit on Wisconsin. With the encoding chosen by inner cross-validation, the VQC exceeded the majority control on Wisconsin and Golub and stayed below classical models on Wisconsin and in several Golub contrasts. No statistically significant difference was detected between the encoding-selected VQC and the classical baselines on Coimbra, or between the Z-map QSVM and PCA-4 classical models; Bayesian equivalence analysis supported practical equivalence only for the QSVM on Wisconsin. On SEER the encoding mattered little: the VQC exceeded the majority control in balanced accuracy (0.548 versus 0.500) but ranked patients worse than logistic regression and random forest (ROC-AUC 0.678 versus 0.745 and 0.733). Full-feature classical models were strongest overall. On hardware (class-balanced test subsets, not natural-prevalence performance), accuracy relative to noise-free simulation changed by −0.071 to +0.048 per run (pooled ibm_fez, Wisconsin: −0.024, 95% CI −0.060 to +0.008), using 5.5 QPU-minutes. Conclusions. Under a fixed training budget, encoding choice was associated with large differences in VQC accuracy on small datasets, so single-encoding negative results should not be generalized; because the Z and ZZ maps also differ in entanglement structure, we do not attribute the gap to encoding alone. No quantum advantage was observed.
Variational quantum classifiers (VQCs) are promising near-term quantum machine-learning models whose performance can depend on both the data-encoding method and circuit noise.This study compared angle encoding, amplitude encoding, and data re-uploading under three levels of simulated depolarizing noise using a fixed four-qubit, three-layer VQC.The comparison was repeated across the Iris versicolor-versus-virginica and Banknote Authentication datasets using multiple stratified data splits and parameter initializations, with a classical multilayer perceptron as a baseline.The relative performance of the encoding methods depended on the dataset: angle encoding produced the highest mean accuracy on Iris, while data re-uploading produced the highest mean accuracy on Banknote Authentication.Although statistical tests identified overall encoding differences in several Banknote conditions, no pairwise comparison remained significant after multiple-comparison correction.The classical baseline outperformed all tested quantum conditions.These findings show that encoding choice can affect VQC classifier behavior under simulated noise, but they do not establish a universally superior encoding method or direct robustness of the encoded quantum states.
An efficient and data-driven encoding scheme is proposed to enhance the performance of variational quantum classifiers. This encoding is specially designed for complex datasets like images and seeks to help the classification task by producing input states that form well-separated clusters in the Hilbert space according to their classification labels. The encoding circuit is trained using a triplet loss function inspired by classical facial recognition algorithms, and class separability is measured via average trace distances between the encoded density matrices. Benchmark tests performed on various binary classification tasks on MNIST and MedMNIST datasets demonstrate considerable improvement over amplitude encoding with the same VQC structure while requiring a much lower circuit depth.