Automated Search for Configurations of Convolutional Neural Network Architectures

Ghamizi, Salah; Cordy, Maxime; Papadakis, Mike; Traon, Yves Le

doi:10.1145/3336294.3336306

Cited by 9 publications

(8 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Both achieve +80% accuracy with Adam Optimizer, data-augmentation and 50 training epochs. The remaining 100 models were produced by FeatureNet [38], a neural architecture search tool that can generate a predefined number of models while maximizing diversity.…”

Section: B Experiments Subjectsmentioning

confidence: 99%

Adversarial Embedding: A robust and elusive Steganography and Watermarking technique

Ghamizi¹,

Cordy²,

Papadakis³

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

We propose adversarial embedding, a new steganography and watermarking technique that embeds secret information within images. The key idea of our method is to use deep neural networks for image classification and adversarial attacks to embed secret information within images. Thus, we use the attacks to embed an encoding of the message within images and the related deep neural network outputs to extract it. The key properties of adversarial attacks (invisible perturbations, nontransferability, resilience to tampering) offer guarantees regarding the confidentiality and the integrity of the hidden messages. We empirically evaluate adversarial embedding using more than 100 models and 1,000 messages. Our results confirm that our embedding passes unnoticed by both humans and steganalysis methods, while at the same time impedes illicit retrieval of the message (less than 13% recovery rate when the interceptor has some knowledge about our model), and is resilient to soft and (to some extent) aggressive image tampering (up to 100% recovery rate under jpeg compression). We further develop our method by proposing a new type of adversarial attack which improves the embedding density (amount of hidden information) of our method to up to 10 bits per pixel.

show abstract

Section: B Experiments Subjectsmentioning

confidence: 99%

Adversarial Embedding: A robust and elusive Steganography and Watermarking technique

Ghamizi¹,

Cordy²,

Papadakis³

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Applications include specifying software product lines [1,20] and predicting systems' performance [18,19]. Conversely, variability management techniques can support the design of machine learning models [8].…”

Section: Introductionmentioning

confidence: 99%

Explicit or Implicit? On Feature Engineering for ML-based Variability-intensive Systems

Temple

Perrouin

2023

Proceedings of the 17th International Working Conference on Variability Modelling of Software-Intensive Systems

View full text Add to dashboard Cite

Software variability engineering benefits from Machine Learning (ML) to learn e.g., variability-aware performance models, explore variants of interest and minimize their energy impact. As the number of applications of combining variability with ML grows, we would like to reflect on what is the core to the configuration process in software variability and inference in ML: feature engineering. These disciplines previously managed features explicitly, easing graceful combinations. Now, deep learning techniques derive automatically obscure but efficient features from data. Shall we give up explicit feature management in variability-intensive systems to embrace machine learning advances?

show abstract

“…In Stein et al., 62 an automated Efficient Global Optimization (EGO) method for exploration of CNNs settings is introduced, where, within image classification, an emphasis is placed on the prediction error minimization. To determine an estimate of the optimal DNN settings, Ghamizi et al 63 . suggested the utilization of Feature Model (FM), where each admissible configuration of the FM leads to an admissible DNN structure.…”

Section: Introductionmentioning

confidence: 99%

An offline multi‐scale unsaturated poromechanics model enabled by self‐designed/self‐improved neural networks

Heider

Suh

Sun

2021

Num Anal Meth Geomechanics

View full text Add to dashboard Cite

Supervised machine learning via artificial neural network (ANN) has gained significant popularity for many geomechanics applications that involves multi‐phase flow and poromechanics. For unsaturated poromechanics problems, the multi‐physics nature and the complexity of the hydraulic laws make it difficult to design the optimal setup, architecture, and hyper‐parameters of the deep neural networks. This paper presents a meta‐modeling approach that utilizes deep reinforcement learning (DRL) to automatically discover optimal neural network settings that maximize a pre‐defined performance metric for the machine learning constitutive laws. This meta‐modeling framework is cast as a Markov Decision Process (MDP) with well‐defined states (subsets of states representing the proposed neural network (NN) settings), actions, and rewards. Following the selection rules, the artificial intelligence (AI) agent, represented in DRL via NN, self‐learns from taking a sequence of actions and receiving feedback signals (rewards) within the selection environment. By utilizing the Monte Carlo Tree Search (MCTS) to update the policy/value networks, the AI agent replaces the human modeler to handle the otherwise time‐consuming trial‐and‐error process that leads to the optimized choices of setup from a high‐dimensional parametric space. This approach is applied to generate two key constitutive laws for the unsaturated poromechanics problems: (1) the path‐dependent retention curve with distinctive wetting and drying paths. (2) The flow in the micropores, governed by an anisotropic permeability tensor. Numerical experiments have shown that the resultant ML‐generated material models can be integrated into a finite element (FE) solver to solve initial‐boundary‐value problems as replacements of the hand‐craft constitutive laws.

show abstract

Automated Search for Configurations of Convolutional Neural Network Architectures

Cited by 9 publications

References 34 publications

Adversarial Embedding: A robust and elusive Steganography and Watermarking technique

Adversarial Embedding: A robust and elusive Steganography and Watermarking technique

Explicit or Implicit? On Feature Engineering for ML-based Variability-intensive Systems

An offline multi‐scale unsaturated poromechanics model enabled by self‐designed/self‐improved neural networks

Contact Info

Product

Resources

About