arXiv subjects

Ljubomir Buturovic

Publications and source records attributed to Ljubomir Buturovic.

4 recordsLinked to original sources

arXiv2026

On Cross-Validation for Hyperparameter Optimization of Deep Learning Image Classifiers

Hyperparameter optimization (HPO) can materially affect the performance of deep learning (DL) image classifiers, but there is little empirical guidance on how to derive the validation signal that drives it, especially for the small sample sizes common in fields such as medical imaging. We compared three HPO protocols in terms of {\em absolute performance-estimation error} (AEE; the absolute difference between the winning configuration's validation AUROC and its test AUROC): fixed holdout (F), reshuffled holdout (R), and 5-fold cross-validation (C). The search space, sampler, training procedure, architecture, and test set were held identical across protocols. We evaluated the protocols on three public datasets spanning two regimes: binary medical imaging (RSNA pneumonia radiographs and binarized HAM10000 skin lesions) and 200-class natural imaging (Tiny ImageNet), across a range of development set sizes $n$ and two backbones (ResNet-18 on all datasets, Vision Transformer (ViT-S/16) on RSNA). On the medical datasets, every point estimate favored cross-validation over both holdout protocols, with reductions in AEE largest at small sample sizes and diminishing as $n$ increased. This pattern remained robust under conservative family-wise adjustment. On Tiny ImageNet, AEE was negligible under all three protocols. Test AUROC was generally similar among protocols. Fixed holdout had lower mean AEE than reshuffled holdout in 11 of 12 medical conditions, although this secondary finding was less uniformly supported. For small-sample medical image classification, we recommend cross-validation-based HPO when computational resources permit because it trades additional computation for a more reliable development-time estimate of subsequent test performance.

cs.CV↗

arXiv2024

Development of Machine Learning Classifiers for Blood-based Diagnosis and Prognosis of Suspected Acute Infections and Sepsis

We applied machine learning to the unmet medical need of rapid and accurate diagnosis and prognosis of acute infections and sepsis in emergency departments. Our solution consists of a Myrna (TM) Instrument and embedded TriVerity (TM) classifiers. The instrument measures abundances of 29 messenger RNAs in patient's blood, subsequently used as features for machine learning. The classifiers convert the input features to an intuitive test report comprising the separate likelihoods of (1) a bacterial infection (2) a viral infection, and (3) severity (need for Intensive Care Unit-level care). In internal validation, the system achieved AUROC = 0.83 on the three-class disease diagnosis (bacterial, viral, or non-infected) and AUROC = 0.77 on binary prognosis of disease severity. The Myrna, TriVerity system was granted breakthrough device designation by the United States Food and Drug Administration (FDA). This engineering manuscript teaches the standard and novel machine learning methods used to translate an academic research concept to a clinical product aimed at improving patient care, and discusses lessons learned.

q-bio.QM↗

arXiv2020

Optimization of Genomic Classifiers for Clinical Deployment: Evaluation of Bayesian Optimization to Select Predictive Models of Acute Infection and In-Hospital Mortality

Acute infection, if not rapidly and accurately detected, can lead to sepsis, organ failure and even death. Current detection of acute infection as well as assessment of a patient's severity of illness are imperfect. Characterization of a patient's immune response by quantifying expression levels of specific genes from blood represents a potentially more timely and precise means of accomplishing both tasks. Machine learning methods provide a platform to leverage this 'host response' for development of deployment-ready classification models. Prioritization of promising classifiers is dependent, in part, on hyperparameter optimization for which a number of approaches including grid search, random sampling and Bayesian optimization have been shown to be effective. We compare HO approaches for the development of diagnostic classifiers of acute infection and in-hospital mortality from gene expression of 29 diagnostic markers. We take a deployment-centered approach to our comprehensive analysis, accounting for heterogeneity in our multi-study patient cohort with our choices of dataset partitioning and hyperparameter optimization objective as well as assessing selected classifiers in external (as well as internal) validation. We find that classifiers selected by Bayesian optimization for in-hospital mortality can outperform those selected by grid search or random sampling. However, in contrast to previous research: 1) Bayesian optimization is not more efficient in selecting classifiers in all instances compared to grid search or random sampling-based methods and 2) we note marginal gains in classifier performance in only specific circumstances when using a common variant of Bayesian optimization (i.e. automatic relevance determination). Our analysis highlights the need for further practical, deployment-centered benchmarking of HO approaches in the healthcare context.

cs.LG↗

arXiv2017

Binary classification models with "Uncertain" predictions

Binary classification models which can assign probabilities to categories such as "the tissue is 75% likely to be tumorous" or "the chemical is 25% likely to be toxic" are well understood statistically, but their utility as an input to decision making is less well explored. We argue that users need to know which is the most probable outcome, how likely that is to be true and, in addition, whether the model is capable enough to provide an answer. It is the last case, where the potential outcomes of the model explicitly include "don't know" that is addressed in this paper. Including this outcome would better separate those predictions that can lead directly to a decision from those where more data is needed. Where models produce an "Uncertain" answer similar to a human reply of "don't know" or "50:50" in the examples we refer to earlier, this would translate to actions such as "operate on tumour" or "remove compound from use" where the models give a "more true than not" answer. Where the models judge the result "Uncertain" the practical decision might be "carry out more detailed laboratory testing of compound" or "commission new tissue analyses". The paper presents several examples where we first analyse the effect of its introduction, then present a methodology for separating "Uncertain" from binary predictions and finally, we provide arguments for its use in practice.

stat.AP↗