An official website of the United States government
Design optimization, and particularly adjoint-based multi-physics shape and topology optimization, is time-consuming and often requires expensive iterations to converge to desired designs. In response, researchers have developed Machine Learning (ML) approaches — often referred to as Inverse Design methods — to either replace or accelerate tools like Topology optimization (TO). However, these methods have their own hidden, non-trivial costs including that of data generation, training, and refinement of ML-produced designs. This begs the question: when is it actually worth learning Inverse Design, compared to just optimizing designs without ML assistance? This paper quantitatively addresses this question by comparing the costs and benefits of three different Inverse Design ML model families on a Topology Optimization (TO) task, compared to just running the optimizer by itself. We explore the relationship between the size of training data and the predictive power of each ML model, as well as the computational and training costs of the models and the extent to which they accelerate or hinder TO convergence. The results demonstrate that simpler models, such as K-Nearest Neighbors and Random Forests, are more effective for TO warmstarting with limited training data, while more complex models, such as Deconvolutional Neural Networks, are preferable with more data. We also emphasize the need to balance the benefits of using larger training sets with the costs of data generation when selecting the appropriate ID model. Finally, the paper addresses some challenges that arise when using ML predictions to warmstart optimization, and provides some suggestions for budget and resource management.
more »
« less
Physics-Guided Hierarchical Neural Networks for Maxwell’s Equations in Plasmonic Metamaterials
While machine learning (ML) has found multiple applications in photonics, traditional “black box” ML models typically require prohibitively large training data sets. Generation of such data, as well as the training processes themselves, consume significant resources, often limiting practical applications of ML. Here, we demonstrate that embedding Maxwell’s equations into ML design and training significantly reduces the required amount of data and improves the physics-consistency and generalizability of ML models, opening the road to practical ML tools that do not need extremely large training sets. The proposed physics-guided machine learning (PGML) approach is illustrated on the example of predicting complex field distributions within hyperbolic metamaterial photonic funnels, based on multilayered plasmonic–dielectric composites. The hierarchical network design used in this study enables knowledge transfer and points to the emergence of effective medium theories within neural network
more »
« less
- PAR ID:
- 10625986
- Publisher / Repository:
- American Chemical Society
- Date Published:
- Journal Name:
- ACS Photonics
- ISSN:
- 2330-4022
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Machine learning (ML) has become a part of the fabric of high-throughput screening and computational discovery of materials. Despite its increasingly central role, challenges remain in fully realizing the promise of ML. This is especially true for the practical acceleration of the engineering of robust materials and the development of design strategies that surpass trial and error or high-throughput screening alone. Depending on the quantity being predicted and the experimental data available, ML can either outperform physics-based models, be used to accelerate such models, or be integrated with them to improve their performance. We cover recent advances in algorithms and in their application that are starting to make inroads toward ( a) the discovery of new materials through large-scale enumerative screening, ( b) the design of materials through identification of rules and principles that govern materials properties, and ( c) the engineering of practical materials by satisfying multiple objectives. We conclude with opportunities for further advancement to realize ML as a widespread tool for practical computational materials design.more » « less
-
Understanding thermal stress evolution in metal additive manufacturing (AM) is crucial for producing high-quality components. Recent advancements in machine learning (ML) have shown great potential for modeling complex multiphysics problems in metal AM. While physics-based simulations face the challenge of high computational costs, conventional data-driven ML models require large, labeled training datasets to achieve accurate predictions. Unfortunately, generating large datasets for ML model training through time-consuming experiments or high-fidelity simulations is highly expensive in metal AM. To address these challenges, this study introduces a physics-informed neural network (PINN) framework that incorporates governing physical laws into deep neural networks (NNs) to predict temperature and thermal stress evolution during the laser metal deposition (LMD) process. The study also discusses enhanced accuracy and efficiency of the PINN model when supplemented with small simulation data. Furthermore, it highlights the PINN transferability, enabling fast predictions with a set of new process parameters using a pre-trained PINN model as an online soft sensor, significantly reducing computation time compared to physics-based numerical models while maintaining accuracy.more » « less
-
We introduce a hybrid model that synergistically combines machine learning (ML) with semiconductor device physics to simulate nanoscale transistors. This approach integrates a physics-based ballistic transistor model with an ML model that predicts ballisticity, enabling flexibility to interface the model with device data. The inclusion of device physics not only enhances the interpretability of the ML model but also streamlines its training process, reducing the necessity for extensive training data. The model's effectiveness is validated on both silicon nanotransistors and carbon nanotube FETs, demonstrating high model accuracy with a simplified ML component. We assess the impacts of various ML models—Multilayer Perceptron (MLP), Recurrent Neural Network (RNN), and RandomForestRegressor (RFR)—on predictive accuracy and training data requirements. Notably, hybrid models incorporating these components can maintain high accuracy with a small training dataset, with the RNN-based model exhibiting better accuracy compared to the MLP and RFR models. The trained hybrid model provides significant speedup compared to device simulations, and can be applied to predict circuit characteristics based on the modeled nanotransistors.more » « less
-
null (Ed.)With the explosion in Big Data, it is often forgotten that much of the data nowadays is generated at the edge. Specifically, a major source of data is users' endpoint devices like phones, smart watches, etc., that are connected to the internet, also known as the Internet-of-Things (IoT). This "edge of data" faces several new challenges related to hardware-constraints, privacy-aware learning, and distributed learning (both training as well as inference). So what systems and machine learning algorithms can we use to generate or exploit data at the edge? Can network science help us solve machine learning (ML) problems? Can IoT-devices help people who live with some form of disability and many others benefit from health monitoring? In this tutorial, we introduce the network science and ML techniques relevant to edge computing, discuss systems for ML (e.g., model compression, quantization, HW/SW co-design, etc.) and ML for systems design (e.g., run-time resource optimization, power management for training and inference on edge devices), and illustrate their impact in addressing concrete IoT applications.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript
- Journal Article:
- https://doi.org/10.1021/acsphotonics.5c00552
