Data Driven VRP: A Neural Network Model to Learn Hidden Preferences for VRP
Adam: метод стохастической оптимизации
2021-01-01
SCID: 54.1/n29teykj
Discuss with AI
DNA-co-polymerized hydrogelscoarse-grained simulationconvolutional neural networkgenetic algorithmmetamorphic material design
Figures from the paper
Abstract (AI)
The traditional Capacitated Vehicle Routing Problem (CVRP) minimizes the total distance of the routes under the capacity constraints of the vehicles. But more often, the objective involves multiple criteria including not only the total distance of the tour but also other factors such as travel costs, travel time, and fuel consumption. Moreover, in reality, there are numerous implicit preferences ingrained in the minds of the route planners and the drivers. Drivers, for instance, have familiarity with certain neighborhoods and knowledge of the state of roads, and often consider the best places for rest and lunch breaks. This knowledge is difficult to formulate and balance when operational routing decisions have to be made. This motivates us to learn the implicit preferences from past solutions and to incorporate these learned preferences in the optimization process. These preferences are in the form of arc probabilities, i.e., the more preferred a route is, the higher is the joint probability. The novelty of this work is the use of a neural network model to estimate the arc probabilities, which allows for additional features and automatic parameter estimation. This first requires identifying suitable features, neural architectures and loss functions, taking into account that there is typically few data available. We investigate the difference with a prior weighted Markov counting approach, and study the applicability of neural networks in this setting.
Key Findings
1
A computational framework combines coarse-grained simulation, convolutional neural network scoring, and genetic algorithms to optimize metamorphic DNA-co-polymerized hydrogel designs.
2
The method autonomously explores a large design space by classifying and scoring simulated hydrogel geometries and evolving batches of candidate designs.
3
The method discovers hydrogel designs capable of transforming into multiple Arabic numerals across different actuation states.
4
The optimization process identifies important material parameters and traits governing hydrogel shape transformation.
5
The study reveals relationships between material scale and the achievable range of shape changes within individual domains, along with trade-offs among design parameters.
Research Object
Metamorphic DNA-co-polymerized hydrogel structures designed to fold into different shapes (Arabic numerals) under different actuation states
Research Subject
design-space exploration and optimization of hydrogel shape-changing behavior, including scale–shape-change relationships and multi-digit transformation
Publication Details
Publication Date
2021-01-01
Journal
Publisher
ISSN
Cited by
9
Open access PDF
Access Type
Author Information
Download PDF
Subscribe to digest
Cited by20
Semi-Supervised Classification with Graph Convolutional Networks2016
Image-to-Image Translation with Conditional Adversarial Networks2017
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale2020
Photo-Realistic Single Image Super-Resolution Using a Generative Adversarial Network2017
Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising2017
mixup: Beyond Empirical Risk Minimization2017
Least Squares Generative Adversarial Networks2017
Google's Neural Machine Translation System: Bridging the Gap between Human and Machine Translation2016
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks2020
Neural Architectures for Named Entity Recognition2016
Heterogeneous Graph Attention Network2019
Attention mechanisms in computer vision: A survey2022
Graph Convolutional Networks for Hyperspectral Image Classification2020
Multimodal Compact Bilinear Pooling for Visual Question Answering and Visual Grounding2016
Automated Breast Ultrasound Lesions Detection Using Convolutional Neural Networks2017
Improving Deep Learning with Generic Data Augmentation2018
HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million\n Narrated Video Clips2019
Interaction Networks for Learning about Objects, Relations and Physics2016
Learning Deep Transformer Models for Machine Translation2019
A Novel Cascade Binary Tagging Framework for Relational Triple Extraction2020