2 code implementations • 9 Apr 2024 • Alexander Chebykin, Peter A. N. Bosman, Tanja Alderliesten
Sharing synthetic medical images is a promising alternative to sharing real images that can improve patient privacy and data security.
no code implementations • 21 Mar 2024 • Arthur Guijt, Dirk Thierens, Tanja Alderliesten, Peter A. N. Bosman
To correct for differing feature representations between these layers we employ stitching, which merges the networks by introducing new layers at crossover points.
no code implementations • 23 Feb 2024 • Monika Grewal, Henrike Westerveld, Peter A. N. Bosman, Tanja Alderliesten
The experiments also show that the proposed MO DIR approach provides a better spread of DIR outputs across the entire trade-off front than simply training multiple neural networks with weights for each objective sampled from a grid of possible values.
no code implementations • 19 Feb 2024 • Damy M. F. Ha, Tanja Alderliesten, Peter A. N. Bosman
The optimal discretization, however, depends on the relations modelled between the variables.
no code implementations • 19 Feb 2024 • Mafalda Malafaia, Thalea Schlender, Peter A. N. Bosman, Tanja Alderliesten
Furthermore, it is important to understand how each modality impacts the final prediction, especially in high-stake domains, so that these models can be used in a trustworthy and responsible manner.
1 code implementation • 16 Feb 2024 • Georgios Andreadis, Tanja Alderliesten, Peter A. N. Bosman
In addition, we propose a new way to model overlapping dependencies in conditional linkage models to maximize the joint sampling of fully interdependent groups of variables.
no code implementations • 15 Feb 2024 • Thalea Schlender, Mafalda Malafaia, Tanja Alderliesten, Peter A. N. Bosman
Experimental results show that both proposed search enhancements have a generally positive impact on the performance of GP-GOMEA, especially when the set of operators to choose from is large and contains higher-arity operators.
no code implementations • 30 Jan 2024 • Georgios Andreadis, Joas I. Mulder, Anton Bouter, Peter A. N. Bosman, Tanja Alderliesten
Although both models have been investigated in detail, a direct comparison has not yet been made, since the models are optimized using very different optimization methods in practice.
1 code implementation • 28 Jul 2023 • Alexander Chebykin, Arkadiy Dushatskiy, Tanja Alderliesten, Peter A. N. Bosman
In this work, we show that simultaneously training and mixing neural networks is a promising way to conduct Neural Architecture Search (NAS).
1 code implementation • 2 Jun 2023 • Arkadiy Dushatskiy, Alexander Chebykin, Tanja Alderliesten, Peter A. N. Bosman
Population Based Training (PBT) is an efficient hyperparameter optimization algorithm.
no code implementations • 27 Mar 2023 • Arthur Guijt, Dirk Thierens, Tanja Alderliesten, Peter A. N. Bosman
Thus, if an asynchronous parallel MBEA is also affected by an evaluation time bias, this could result in learned models to be less suited to solving the problem, reducing performance.
no code implementations • 20 Mar 2023 • Vangelis Kostoulas, Peter A. N. Bosman, Tanja Alderliesten
The results show that small improvements in metrics can be achieved by advancing and merging architectures, but the predictions of the models are quite similar (most models achieve on average more than 0. 8 Dice Coefficient when compared to the outputs of other models).
no code implementations • 8 Mar 2023 • Georgios Andreadis, Peter A. N. Bosman, Tanja Alderliesten
A recent multi-objective approach that uses the Multi-Objective Real-Valued Gene-pool Optimal Mixing Evolutionary Algorithm (MO-RV-GOMEA) and a dual-dynamic mesh transformation model has shown promise, exposing the trade-offs inherent to image registration problems and modeling large deformations in 2D.
no code implementations • 21 Feb 2023 • Monika Grewal, Dustin van Weersel, Henrike Westerveld, Peter A. N. Bosman, Tanja Alderliesten
Following this motivation, we present an approach to learn a deep learning model for the automatic segmentation of Organs at Risk (OARs) in cervical cancer radiation treatment from a large clinically available dataset of Computed Tomography (CT) scans containing data inhomogeneity, label noise, and missing annotations.
no code implementations • 5 Apr 2022 • Marco Virgolin, Eric Medvet, Tanja Alderliesten, Peter A. N. Bosman
Interpretability can be critical for the safe and responsible use of machine learning models in high-stakes applications.
no code implementations • 17 Mar 2022 • Renzo J. Scholman, Anton Bouter, Leah R. M. Dickhoff, Tanja Alderliesten, Peter A. N. Bosman
Even if a Multi-modal Multi-Objective Evolutionary Algorithm (MMOEA) is designed to find solutions well spread over all locally optimal approximation sets of a Multi-modal Multi-objective Optimization Problem (MMOP), there is a risk that the found set of solutions is not smoothly navigable because the solutions belong to various niches, reducing the insight for decision makers.
no code implementations • 16 Mar 2022 • Leah R. M. Dickhoff, Ellen M. Kerkhof, Heloisa H. Deuzeman, Carien L. Creutzberg, Tanja Alderliesten, Peter A. N. Bosman
For this reason, we propose a novel adaptive objective configuration method to use with MO-RV-GOMEA so that we can accommodate additional aims of this nature.
1 code implementation • 11 Mar 2022 • Arthur Guijt, Dirk Thierens, Tanja Alderliesten, Peter A. N. Bosman
This cannot be modelled sufficiently well when using linkage models that aim at capturing a single type of linkage structure, deteriorating the advantages brought by MBEAs.
1 code implementation • 8 Mar 2022 • Alexander Chebykin, Tanja Alderliesten, Peter A. N. Bosman
A recent method called Neural Architecture Transfer (NAT) further improves the efficiency of NAS for computer vision tasks by using a multi-objective evolutionary algorithm to find high-quality subnetworks of a supernetwork pretrained on ImageNet.
1 code implementation • 1 Mar 2022 • Thomas Uriot, Marco Virgolin, Tanja Alderliesten, Peter Bosman
We find that various GP methods can be competitive with state-of-the-art DR algorithms and that they have the potential to produce interpretable DR mappings.
no code implementations • 24 Feb 2022 • Arkadiy Dushatskiy, Gerry Lowe, Peter A. N. Bosman, Tanja Alderliesten
In experiments with a real clinical dataset of CT scans with prostate segmentations, our approach provides an improvement of several percentage points in terms of Dice and surface Dice coefficients compared to when all network paths are trained on all training data.
no code implementations • 23 Feb 2022 • Martijn M. A. Bosma, Arkadiy Dushatskiy, Monika Grewal, Tanja Alderliesten, Peter A. N. Bosman
The design of the best possible medical image segmentation DNNs, however, is task-specific.
no code implementations • 22 Feb 2022 • Georgios Andreadis, Peter A. N. Bosman, Tanja Alderliesten
Concordantly, this work introduces the first method for multi-objective 3D deformable image registration, using a 3D dual-dynamic grid transformation model based on simplex meshes while still supporting the incorporation of annotated guidance information and multi-resolution schemes.
no code implementations • 14 Feb 2022 • Dazhuang Liu, Marco Virgolin, Tanja Alderliesten, Peter A. N. Bosman
Genetic programming (GP) is one of the best approaches today to discover symbolic regression models.
no code implementations • 10 Feb 2022 • Marco Virgolin, Andrea De Lorenzo, Tanja Alderliesten, Peter A. N. Bosman
Our results indicate that adult data can be considered to be a meaningful augmentation of pediatric data for the recognition of emotional facial expression in children, and open up the possibility for other applications of contrastive learning to improve pediatric care by complementing data of children with that of adults.
no code implementations • 4 Feb 2022 • Arkadiy Dushatskiy, Tanja Alderliesten, Peter A. N. Bosman
Due to stochastic factors in neural network initialization, training, and the chosen train/validation dataset split, the performance evaluation of a neural network architecture, which is often based on a single learning run, is also stochastic.
no code implementations • 6 Sep 2021 • Monika Grewal, Jan Wiersma, Henrike Westerveld, Peter A. N. Bosman, Tanja Alderliesten
Conclusions: DCNN-Match learns to predict landmark correspondences in 3D medical images in a self-supervised manner, which can improve DIR performance.
1 code implementation • 16 Apr 2021 • Arkadiy Dushatskiy, Tanja Alderliesten, Peter A. N. Bosman
We propose a novel surrogate-assisted Evolutionary Algorithm for solving expensive combinatorial optimization problems.
1 code implementation • 8 Feb 2021 • Timo M. Deist, Monika Grewal, Frank J. W. M. Dankers, Tanja Alderliesten, Peter A. N. Bosman
We discuss and illustrate why training processes to approximate Pareto fronts need to optimize on fronts of individual training samples instead of on only the front of average losses.
3 code implementations • 9 Jul 2020 • Timo M. Deist, Stefanus C. Maree, Tanja Alderliesten, Peter A. N. Bosman
On several bi-objective benchmarks, we find that gradient-based algorithms outperform the tested EAs by obtaining a better hypervolume with fewer evaluations whenever exact gradients of the multiple objective functions are available and in case of small evaluation budgets.
Optimization and Control
no code implementations • 23 Jan 2020 • Arkadiy Dushatskiy, Adriënne M. Mendrik, Peter A. N. Bosman, Tanja Alderliesten
There has recently been great progress in automatic segmentation of medical images with deep learning algorithms.
2 code implementations • 21 Jan 2020 • Monika Grewal, Timo M. Deist, Jan Wiersma, Peter A. N. Bosman, Tanja Alderliesten
We tested the approach on 22, 206 pairs of 2D slices with varying levels of intensity, affine, and elastic transformations.
no code implementations • 9 Sep 2019 • Marco Virgolin, Ziyuan Wang, Tanja Alderliesten, Peter A. N. Bosman
To assess the effects of radiation therapy, treatment plans are typically simulated on phantoms, i. e., virtual surrogates of patient anatomy.
no code implementations • 4 Jul 2019 • Marco Virgolin, Tanja Alderliesten, Peter A. N. Bosman
In this article, we assess to what extent GP still performs favorably at feature construction when constructing features that are (1) Of small-enough number, to enable visualization of the behavior of the ML model; (2) Of small-enough size, to enable interpretability of the features themselves; (3) Of sufficient informative power, to retain or even improve the performance of the ML algorithm.
1 code implementation • 3 Apr 2019 • Marco Virgolin, Tanja Alderliesten, Cees Witteveen, Peter A. N. Bosman
We show that the non-uniformity in the distribution of the genotype in GP populations negatively biases LL, and propose a method to correct for this.