SHSharon Yalov Handzel

Sharon Yalov Handzel

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Sharon Yalov Handzel has a total of 8 co-authors. Here are the top co-authors by number of shared publications:

  • Brian Rikshpun

    Department of Machine Learning and AI Afeka Tel‐Aviv Academic College of Engineering Tel‐Aviv Israel

    1 shared publication
  • Chava Kimchi-Sarfaty

    Hemostasis Branch 1, Division of Hemostasis, Office of Plasma Protein Therapeutics, Office of Therapeutic Products, Center for Biologics Evaluation & Research, US Food and Drug Administration, Silver Spring, USA

    1 shared publication
  • Din Kosberg

    Afeka College of Engineering

    1 shared publication
  • Ido Cohen

    1 shared publication
  • Keren Glickman

    Afeka Tel-Aviv Academic College of Engineering, Tel Aviv, Israel

    1 shared publication
  • Nathan Clement

    Hemostasis Branch 1, Division of Hemostasis, Office of Plasma Protein Therapeutics CMC, Office Therapeutic Products, Center for Biologics Evaluation and Research, US Food and Drug Administration Silver Spring Maryland USA

    1 shared publication

Latest Publications

  • Open Access

    Longitudinal Component Level Analysis of Neural Network Training Dynamics Beyond Accuracy and Loss

    by Sharon Yalov Handzel, Din Kosberg

    Accuracy and loss summarize predictive performance but provide limited information about how individual components evolve during training. This study presents a non-intrusive observation framework that records epoch-level descriptors of neuron activations, connection weights, gradients, updates, and biases without changing the training objective or optimizer. The framework was evaluated in five controlled experiments on multilayer perceptrons trained on MNIST and the Wisconsin Diagnostic Breast Cancer dataset. The final protocol comprised five random seeds, 30 MNIST epochs, 100 WDBC epochs, and 130 dataset-condition-seed combinations examining repeated runs, network width, learning rate, activation function, and descriptor-guided pruning. Models with similar validation accuracy frequently occupied different internal regimes. On MNIST, the coefficient of variation of gradient magnitude across seeds was 0.239, compared with 0.0021 for accuracy. Increasing learning rate from 0.0001 to 0.01 increased persistent inactivity from 0.0052 to 0.3656, while four activation functions separated by less than one percentage point of accuracy exhibited distinct inactivity and saturation trajectories. In the functional evaluation, descriptor-guided edge pruning caused less accuracy loss than layer-matched random pruning in all 15 MNIST seed-fraction comparisons and in 9 of 15 WDBC comparisons. However, the composite score and weight-magnitude pruning performed comparably at 5% to 20% sparsity. The findings show that component-level temporal descriptors reveal localized and reversible processes that are obscured by final global metrics. They support the use of internal dynamics as a complementary diagnostic layer, but do not establish a universal predictor of model quality or a pruning criterion superior to magnitude.

  • Open Access

    An Alignment Free Framework for Taxonomic Inference From Codon and Codon‐Pair Usage

    by Sharon Yalov Handzel, Brian Rikshpun, Nigam Padhiar, Nathan Clement, Chava Kimchi‐Sarfaty

    Alignment‐free signals in coding sequences provide a scalable route to taxonomic inference, quality control of large phylogenies, and rapid screening of genomic data. We present a model‐agnostic framework that represents genomes or coding‐sequence collections using codon usage (64‐D) and codon‐pair usage (4096‐D) profiles, and we introduce Taxonomic Consistency (TC), a simple, rank‐aware external index, for evaluating supervised predictions or unsupervised clustering against the hierarchical taxonomy. Across multiple taxonomic ranks (e.g., Domain, Order, etc.), compact supervised models and standard clustering methods are assessed with both internal (Silhouette) and external (TC) validation. In large‐scale experiments, handling class imbalance and applying principled normalization had a greater impact on performance than sequence‐level preprocessing, and codon usage profiles yielded the highest TC and coherent unsupervised structure, while codon‐pair features provided complementary resolution within specific clades. We release open, versioned code and reproducible workflows to facilitate adoption in ecological and evolutionary pipelines. Taken together, alignment‐free codon and codon‐pair representations, paired with TC for transparent external validation, offer a practical complement to tree‐based methods—especially for rapid screening, sanity checks, and exploratory analyses at higher ranks.

  • Open Access

    The impact of synthetic data rates in imbalanced datasets on convergence characteristics of deep learning networks

    by Sharon Yalov-Handzel, Keren Glickman

    The application of machine learning to imbalanced datasets may cause undesirably biased results. Some preprocessing applied to the dataset may avoid distortion against the minority class. A common approach for implementation of such preprocessing is by data synthesis. In this study, we compare the impact of varying rates of synthetic data assimilated into the original dataset on the training performance of a neural network (NN). The synthetic data was generated by two different algorithms: Conditional Generative Adversarial Network (CTGAN) and Triple Based Variational Autoencoder (TVAE), an encoder-based data generator. Varying rates of synthetic data were assimilated into two different medical datasets where the attribute of the patient’s age is imbalanced. In the first, a birth deliveries dataset, the NN was implemented to solve a regression problem; the second dataset, containing information about Covid-19 patients, was used to solve a classification problem. From these two original datasets, an additional 26 datasets were derived, with varying rates of synthetic records that were generated by two different algorithms. The training performance of a NN on these datasets was compared in terms of accuracy, convergence speed, and two novel metrics designed to quantify internal oscillations of the network’s weights during training. The results demonstrate that both the data synthesis method and the proportion of synthetic data significantly affect model accuracy and training dynamics. This study contributes a novel framework for assessing convergence stability under imbalanced conditions using deep generative synthetic data.