Neural Architecture Search with Reinforcement Learning

Abstract

Neural networks are powerful and flexible models that work well for many difficult learning tasks in image, speech and natural language understanding. Despite their success, neural networks are still hard to design. In this paper, we use a recurrent network to generate the model descriptions of neural networks and train this RNN with reinforcement learning to maximize the expected accuracy of the generated architectures on a validation set. On the CIFAR-10 dataset, our method, starting from scratch, can design a novel network architecture that rivals the best human-invented architecture in terms of test set accuracy. Our CIFAR-10 model achieves a test error rate of 3.65, which is 0.09 percent better and 1.05x faster than the previous state-of-the-art model that used a similar architectural scheme. On the Penn Treebank dataset, our model can compose a novel recurrent cell that outperforms the widely-used LSTM cell, and other state-of-the-art baselines. Our cell achieves a test set perplexity of 62.4 on the Penn Treebank, which is 3.6 perplexity better than the previous state-of-the-art model. The cell can also be transferred to the character language modeling task on PTB and achieves a state-of-the-art perplexity of 1.214.

References in corpus (6)

• Very Deep Convolutional Networks for Large-Scale Image Recognition
• Striving for Simplicity: The All Convolutional Net
• Recurrent Neural Network Regularization
• Theoretical Models of Learning to Learn
• Pointer Sentinel Mixture Models
• Tying Word Vectors and Word Classifiers: A Loss Framework for Language Modeling

Cited by in corpus (389)

• A Brief Survey of Deep Reinforcement Learning
• Searching for Activation Functions
• Regularizing and Optimizing LSTM Language Models
• Multi-Task Learning with Deep Neural Networks: A Survey
• SMASH: One-Shot Model Architecture Search through HyperNetworks
• Robust Adversarial Reinforcement Learning
• Learning to reinforcement learn
• NAS-FAS: Static-Dynamic Central Difference Network Search for Face Anti-Spoofing
• Hierarchical Neural Architecture Search for Deep Stereo Matching
• Neural Optimizer Search with Reinforcement Learning
• A Survey on Neural Speech Synthesis
• Simple And Efficient Architecture Search for Convolutional Neural Networks
• Tiny Machine Learning: Progress and Futures
• Automated Machine Learning: State-of-The-Art and Open Challenges
• AutoSlim: Towards One-Shot Architecture Search for Channel Numbers
• Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications
• Optimization for deep learning: theory and algorithms
• Neural Machine Translation and Sequence-to-sequence Models: A Tutorial
• N2N Learning: Network to Network Compression via Policy Gradient Reinforcement Learning
• FasterSeg: Searching for Faster Real-time Semantic Segmentation
• AdaNet: Adaptive Structural Learning of Artificial Neural Networks
• Automated Machine Learning in Practice: State of the Art and Recent Results
• Edge Intelligence: Paving the Last Mile of Artificial Intelligence with Edge Computing
• PP-LCNet: A Lightweight CPU Convolutional Neural Network
• Mind Mappings: Enabling Efficient Algorithm-Accelerator Mapping Space Search
• Recent Progress in the CUHK Dysarthric Speech Recognition System
• Adversarial AutoAugment
• Partial Connection Based on Channel Attention for Differentiable Neural Architecture Search
• Peephole: Predicting Network Performance Before Training
• SpArSe: Sparse Architecture Search for CNNs on Resource-Constrained Microcontrollers
• Efficient Hyperparameter Optimization in Deep Learning Using a Variable Length Genetic Algorithm
• Exploring Randomly Wired Neural Networks for Image Recognition
• Learn to Grow: A Continual Structure Learning Framework for Overcoming Catastrophic Forgetting
• Probabilistic Neural Architecture Search
• Learning to Continually Learn
• Resurrecting the sigmoid in deep learning through dynamical isometry: theory and practice
• Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement
• Dynamic Evaluation of Neural Sequence Models
• On Neural Architecture Search for Resource-Constrained Hardware Platforms
• Cross-dimensional transfer learning in medical image segmentation with deep learning
• Out of Distribution Generalization in Machine Learning
• Image quality assessment for machine learning tasks using meta-reinforcement learning
• Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective
• LegoDNN: Block-grained Scaling of Deep Neural Networks for Mobile Vision
• On the State of the Art of Evaluation in Neural Language Models
• Generative Teaching Networks: Accelerating Neural Architecture Search by Learning to Generate Synthetic Training Data
• Lightweight Neural Architecture Search for Temporal Convolutional Networks at the Edge
• RepVGG: Making VGG-style ConvNets Great Again
• Hybrid Tensor Decomposition in Neural Network Compression
• A Hybrid Convolutional Variational Autoencoder for Text Generation
• A neural network walks into a lab: towards using deep nets as models for human behavior
• Revisiting Activation Regularization for Language RNNs
• A Comprehensive Review of Modern Object Segmentation Approaches
• DeepFD: Automated Fault Diagnosis and Localization for Deep Learning Programs
• Fast-Slow Recurrent Neural Networks
• Comparison and Benchmarking of AI Models and Frameworks on Mobile Devices
• AGAN: Towards Automated Design of Generative Adversarial Networks
• Multi-Objective Reinforced Evolution in Mobile Neural Architecture Search
• Quantum Architecture Search via Deep Reinforcement Learning
• RMM: Reinforced Memory Management for Class-Incremental Learning
• Hyp-RL : Hyperparameter Optimization by Reinforcement Learning
• Deeper Insights into Weight Sharing in Neural Architecture Search
• MaxUp: A Simple Way to Improve Generalization of Neural Network Training
• PSLT: A Light-weight Vision Transformer with Ladder Self-Attention and Progressive Shift
• StressNAS: Affect State and Stress Detection Using Neural Architecture Search
• Designing the Topology of Graph Neural Networks: A Novel Feature Fusion Perspective
• Neural Graph Evolution: Towards Efficient Automatic Robot Design
• Blending Diverse Physical Priors with Neural Networks
• RLCard: A Toolkit for Reinforcement Learning in Card Games
• Computation Reallocation for Object Detection
• FBNetV2: Differentiable Neural Architecture Search for Spatial and Channel Dimensions
• Recurrent Additive Networks
• On Training Sample Memorization: Lessons from Benchmarking Generative Modeling with a Large-scale Competition
• One Proxy Device Is Enough for Hardware-Aware Neural Architecture Search
• Learning Graph Convolutional Network for Skeleton-based Human Action Recognition by Neural Searching
• Learning to Prune Filters in Convolutional Neural Networks
• RC-DARTS: Resource Constrained Differentiable Architecture Search
• Connectivity Learning in Multi-Branch Networks
• PNAS-MOT: Multi-Modal Object Tracking with Pareto Neural Architecture Search
• DropNAS: Grouped Operation Dropout for Differentiable Architecture Search
• Enhancing Neural Architecture Search with Multiple Hardware Constraints for Deep Learning Model Deployment on Tiny IoT Devices
• Diverse Branch Block: Building a Convolution as an Inception-like Unit
• Intriguing Properties of Adversarial Examples
• Deep Long Audio Inpainting
• A Comprehensive Survey of Multilingual Neural Machine Translation
• XNAS: Neural Architecture Search with Expert Advice
• APQ: Joint Search for Network Architecture, Pruning and Quantization Policy
• BioNetExplorer: Architecture-Space Exploration of Bio-Signal Processing Deep Neural Networks for Wearables
• AutoEmb: Automated Embedding Dimensionality Search in Streaming Recommendations
• Not All Ops Are Created Equal!
• PSViT: Better Vision Transformer via Token Pooling and Attention Sharing
• Reinforcement Learning Driven Heuristic Optimization
• Few-Shot Speaker Identification Using Lightweight Prototypical Network with Feature Grouping and Interaction
• HARK Side of Deep Learning -- From Grad Student Descent to Automated Machine Learning
• HardCoRe-NAS: Hard Constrained diffeRentiable Neural Architecture Search
• Question Answering from Unstructured Text by Retrieval and Comprehension
• LPYOLO: Low Precision YOLO for Face Detection on FPGA
• MoViNets: Mobile Video Networks for Efficient Video Recognition
• Chameleon: Adaptive Code Optimization for Expedited Deep Neural Network Compilation
• Learned Low Precision Graph Neural Networks
• Neural Feature Search for RGB-Infrared Person Re-Identification
• Exploring Benefits of Transfer Learning in Neural Machine Translation
• Real-time Federated Evolutionary Neural Architecture Search
• Hardware-Centric AutoML for Mixed-Precision Quantization
• Automatic Mixed-Precision Quantization Search of BERT
• Multigrid Predictive Filter Flow for Unsupervised Learning on Videos
• DMCP: Differentiable Markov Channel Pruning for Neural Networks
• Using Small Proxy Datasets to Accelerate Hyperparameter Search
• Large Scale Evolution of Convolutional Neural Networks Using Volunteer Computing
• Funnel Activation for Visual Recognition
• AutoSpeech: Neural Architecture Search for Speaker Recognition
• Evolutionary Architecture Search for Graph Neural Networks
• Efficient Differentiable Neural Architecture Search with Meta Kernels
• Towards NNGP-guided Neural Architecture Search
• Convolution Neural Network Architecture Learning for Remote Sensing Scene Classification
• Learning to Optimize in Swarms
• AutoPose: Searching Multi-Scale Branch Aggregation for Pose Estimation
• Fine-Grained Neural Architecture Search
• Learnable Embedding Space for Efficient Neural Architecture Compression
• FPGA/DNN Co-Design: An Efficient Design Methodology for IoT Intelligence on the Edge
• SalNAS: Efficient Saliency-prediction Neural Architecture Search with self-knowledge distillation
• Reinforcement Learning and Adaptive Sampling for Optimized DNN Compilation
• Balanced Symmetric Cross Entropy for Large Scale Imbalanced and Noisy Data
• Dynamic Optimization of Neural Network Structures Using Probabilistic Modeling
• AttentiveNAS: Improving Neural Architecture Search via Attentive Sampling
• Pruning In Time (PIT): A Lightweight Network Architecture Optimizer for Temporal Convolutional Networks
• BETANAS: BalancEd TrAining and selective drop for Neural Architecture Search
• Multiresolution Convolutional Autoencoders
• Gradient-free Policy Architecture Search and Adaptation
• SapientML: Synthesizing Machine Learning Pipelines by Learning from Human-Written Solutions
• You Only Compress Once: Towards Effective and Elastic BERT Compression via Exploit-Explore Stochastic Nature Gradient
• DVOLVER: Efficient Pareto-Optimal Neural Network Architecture Search
• Bayesian Learning of Neural Network Architectures
• A Review of Meta-Reinforcement Learning for Deep Neural Networks Architecture Search
• Core-set Sampling for Efficient Neural Architecture Search
• Neuroevolution of Neural Network Architectures Using CoDeepNEAT and Keras
• PAMS: Quantized Super-Resolution via Parameterized Max Scale
• AutoGCN -- Towards Generic Human Activity Recognition with Neural Architecture Search
• Map-based Multi-Policy Reinforcement Learning: Enhancing Adaptability of Robots by Deep Reinforcement Learning
• From Federated Learning to Federated Neural Architecture Search: A Survey
• Meta-Learning for Contextual Bandit Exploration
• Techniques for Automated Machine Learning
• Delta-STN: Efficient Bilevel Optimization for Neural Networks using Structured Response Jacobians
• Towards Similarity Graphs Constructed by Deep Reinforcement Learning
• Neural Architecture Optimization with Graph VAE
• Improving One-shot NAS by Suppressing the Posterior Fading
• DSRNA: Differentiable Search of Robust Neural Architectures
• ProBO: Versatile Bayesian Optimization Using Any Probabilistic Programming Language
• Improving Neural Architecture Search Image Classifiers via Ensemble Learning
• Automatic Model Selection for Neural Networks
• A Generalized Zero-Shot Quantization of Deep Convolutional Neural Networks via Learned Weights Statistics
• DartsReNet: Exploring new RNN cells in ReNet architectures
• Nonparametric Neural Networks
• An Introduction to Neural Architecture Search for Convolutional Networks
• Deep Networks from the Principle of Rate Reduction
• Improved Conformer-based End-to-End Speech Recognition Using Neural Architecture Search
• Learning to update Auto-associative Memory in Recurrent Neural Networks for Improving Sequence Memorization
• Exploration of Quantum Neural Architecture by Mixing Quantum Neuron Designs
• Multi-objective Asynchronous Successive Halving
• Deep RGB-D Saliency Detection with Depth-Sensitive Attention and Automatic Multi-Modal Fusion
• A Novel Framework for Neural Architecture Search in the Hill Climbing Domain
• Data-driven Neural Architecture Learning For Financial Time-series Forecasting
• Measuring Visual Generalization in Continuous Control from Pixels
• Neural Architecture Search For Fault Diagnosis
• Learning Versatile Convolution Filters for Efficient Visual Recognition
• Automatically Searching for U-Net Image Translator Architecture
• Powering One-shot Topological NAS with Stabilized Share-parameter Proxy
• Full-Cycle Energy Consumption Benchmark for Low-Carbon Computer Vision
• Constrained deep neural network architecture search for IoT devices accounting hardware calibration
• Accuracy vs. Efficiency: Achieving Both through FPGA-Implementation Aware Neural Architecture Search
• Joint Search of Data Augmentation Policies and Network Architectures
• S3NAS: Fast NPU-aware Neural Architecture Search Methodology
• PV-NAS: Practical Neural Architecture Search for Video Recognition
• NAX: Co-Designing Neural Network and Hardware Architecture for Memristive Xbar based Computing Systems
• Semantic-Based Neural Network Repair
• Domain-Aware Dynamic Networks
• On the Privacy Risks of Cell-Based NAS Architectures
• Dancing along Battery: Enabling Transformer with Run-time Reconfigurability on Mobile Devices
• K-shot NAS: Learnable Weight-Sharing for NAS with K-shot Supernets
• TransTailor: Pruning the Pre-trained Model for Improved Transfer Learning
• POPNASv2: An Efficient Multi-Objective Neural Architecture Search Technique
• Multi-objective Neural Architecture Search via Non-stationary Policy Gradient
• Exploiting Uncertainties from Ensemble Learners to Improve Decision-Making in Healthcare AI
• Latency-Aware Neural Architecture Search with Multi-Objective Bayesian Optimization
• Transfer Learning to Learn with Multitask Neural Model Search
• DARC: Differentiable ARchitecture Compression
• Deep Neural Architecture Search with Deep Graph Bayesian Optimization
• Catalyst.RL: A Distributed Framework for Reproducible RL Research
• Towards Oracle Knowledge Distillation with Neural Architecture Search
• How can AI Automate End-to-End Data Science?
• Characterizing Deep Learning Training Workloads on Alibaba-PAI
• Synthetic Sample Selection via Reinforcement Learning
• SOTERIA: In Search of Efficient Neural Networks for Private Inference
• Policy-GNN: Aggregation Optimization for Graph Neural Networks
• MixSearch: Searching for Domain Generalized Medical Image Segmentation Architectures
• Self-Organizing Intelligent Matter: A blueprint for an AI generating algorithm
• Discovering Robust Convolutional Architecture at Targeted Capacity: A Multi-Shot Approach
• FNAS: Uncertainty-Aware Fast Neural Architecture Search
• GDOD: Effective Gradient Descent using Orthogonal Decomposition for Multi-Task Learning
• RSO: A Gradient Free Sampling Based Approach For Training Deep Neural Networks
• Synthetic Petri Dish: A Novel Surrogate Model for Rapid Architecture Search
• Layer Folding: Neural Network Depth Reduction using Activation Linearization
• Fine-Grained Stochastic Architecture Search
• Efficient Architecture Search for Continual Learning
• MobileDepth: Efficient Monocular Depth Prediction on Mobile Devices
• Self-Assembling Modular Networks for Interpretable Multi-Hop Reasoning
• SAIA: Split Artificial Intelligence Architecture for Mobile Healthcare System
• A Learning-Based Tune-Free Control Framework for Large Scale Autonomous Driving System Deployment
• Privacy-preserving Collaborative Learning with Automatic Transformation Search
• Learn Basic Skills and Reuse: Modularized Adaptive Neural Architecture Search (MANAS)
• Dynamically Throttleable Neural Networks (TNN)
• Generalizable Resource Allocation in Stream Processing via Deep Reinforcement Learning
• Challenges for cognitive decoding using deep learning methods
• HOLMES: Health OnLine Model Ensemble Serving for Deep Learning Models in Intensive Care Units
• ConvNets vs. Transformers: Whose Visual Representations are More Transferable?
• DC-NAS: Divide-and-Conquer Neural Architecture Search
• Latent Domain Learning with Dynamic Residual Adapters
• Rethinking Performance Estimation in Neural Architecture Search
• OnlineAugment: Online Data Augmentation with Less Domain Knowledge
• Binarized Neural Architecture Search for Efficient Object Recognition
• Evolving Deep Neural Networks by Multi-objective Particle Swarm Optimization for Image Classification
• Video Action Recognition Via Neural Architecture Searching
• HyperSTAR: Task-Aware Hyperparameters for Deep Networks
• TextNAS: A Neural Architecture Search Space tailored for Text Representation
• Learning Graph Representation of Person-specific Cognitive Processes from Audio-visual Behaviours for Automatic Personality Recognition
• RAPDARTS: Resource-Aware Progressive Differentiable Architecture Search
• Parameter Prediction for Unseen Deep Architectures
• ACDC: Weight Sharing in Atom-Coefficient Decomposed Convolution
• A Generative Model for Sampling High-Performance and Diverse Weights for Neural Networks
• Intelligence, physics and information -- the tradeoff between accuracy and simplicity in machine learning
• Exploiting Operation Importance for Differentiable Neural Architecture Search
• Data-Driven Compression of Convolutional Neural Networks
• StyleNAS: An Empirical Study of Neural Architecture Search to Uncover Surprisingly Fast End-to-End Universal Style Transfer Networks
• Transfer Learning for Multi-lingual Tasks -- a Survey
• WeNet: Weighted Networks for Recurrent Network Architecture Search
• Neural Architecture Search for Gliomas Segmentation on Multimodal Magnetic Resonance Imaging
• Regularize, Expand and Compress: Multi-task based Lifelong Learning via NonExpansive AutoML
• Capacity allocation analysis of neural networks: A tool for principled architecture design
• Single Person Pose Estimation: A Survey
• AntMan: Sparse Low-Rank Compression to Accelerate RNN inference
• Learning Less-Overlapping Representations
• CompOFA: Compound Once-For-All Networks for Faster Multi-Platform Deployment
• NAS-HPO-Bench-II: A Benchmark Dataset on Joint Optimization of Convolutional Neural Network Architecture and Training Hyperparameters
• Learned Indexes for Dynamic Workloads
• ResNetX: a more disordered and deeper network architecture
• Effective, Efficient and Robust Neural Architecture Search
• Operation-level Progressive Differentiable Architecture Search
• EPNAS: Efficient Progressive Neural Architecture Search
• Multi-Pass Transformer for Machine Translation
• AdaDeep: A Usage-Driven, Automated Deep Model Compression Framework for Enabling Ubiquitous Intelligent Mobiles
• Piecewise Linear Units Improve Deep Neural Networks
• AutoAdapt: Automated Segmentation Network Search for Unsupervised Domain Adaptation
• RandomNet: Towards Fully Automatic Neural Architecture Design for Multimodal Learning
• Binarized Neural Architecture Search
• Auptimizer -- an Extensible, Open-Source Framework for Hyperparameter Tuning
• Improving Binary Neural Networks through Fully Utilizing Latent Weights
• NeuroDB: A Neural Network Framework for Answering Range Aggregate Queries and Beyond
• A Feasible Framework for Arbitrary-Shaped Scene Text Recognition
• Efficient Neural Architecture Search with Performance Prediction
• Differentiable NAS Framework and Application to Ads CTR Prediction
• Automated Problem Identification: Regression vs Classification via Evolutionary Deep Networks
• SuperShaper: Task-Agnostic Super Pre-training of BERT Models with Variable Hidden Dimensions
• Explaining Transition Systems through Program Induction
• ImmuNetNAS: An Immune-network approach for searching Convolutional Neural Network Architectures
• Expandable YOLO: 3D Object Detection from RGB-D Images
• Robustifying DARTS by Eliminating Information Bypass Leakage via Explicit Sparse Regularization
• RicciNets: Curvature-guided Pruning of High-performance Neural Networks Using Ricci Flow
• Genetic Network Architecture Search
• Learning to Cope with Adversarial Attacks
• DeepSwarm: Optimising Convolutional Neural Networks using Swarm Intelligence
• NAS-TC: Neural Architecture Search on Temporal Convolutions for Complex Action Recognition
• Yet Another Intermediate-Level Attack
• On the Orthogonality of Knowledge Distillation with Other Techniques: From an Ensemble Perspective
• Tuning a variational autoencoder for data accountability problem in the Mars Science Laboratory ground data system
• ENAS4D: Efficient Multi-stage CNN Architecture Search for Dynamic Inference
• Rotational Unit of Memory
• Multi-objective Search of Robust Neural Architectures against Multiple Types of Adversarial Attacks
• GEVO: GPU Code Optimization using Evolutionary Computation
• In folly ripe. In reason rotten. Putting machine theology to rest
• Memory-Efficient Differentiable Transformer Architecture Search
• Single-level Optimization For Differential Architecture Search
• Model-Agnostic Meta-Learning using Runge-Kutta Methods
• Using Program Induction to Interpret Transition System Dynamics
• Sample Efficient Ensemble Learning with Catalyst.RL
• URNet : User-Resizable Residual Networks with Conditional Gating Module
• Compositional Models: Multi-Task Learning and Knowledge Transfer with Modular Networks
• TimeGate: Conditional Gating of Segments in Long-range Activities
• HAO: Hardware-aware neural Architecture Optimization for Efficient Inference
• FP-NAS: Fast Probabilistic Neural Architecture Search
• Learning image quality assessment by reinforcing task amenable data selection
• Exploiting Syntactic Structure for Better Language Modeling: A Syntactic Distance Approach
• Compact CNN Structure Learning by Knowledge Distillation
• Software/Hardware Co-design for Multi-modal Multi-task Learning in Autonomous Systems
• Evolutionary Algorithm Enhanced Neural Architecture Search for Text-Independent Speaker Verification
• Pretraining Neural Architecture Search Controllers with Locality-based Self-Supervised Learning
• Efficient Model Performance Estimation via Feature Histories
• Multi-level Feature Fusion-based CNN for Local Climate Zone Classification from Sentinel-2 Images: Benchmark Results on the So2Sat LCZ42 Dataset
• Self-Reorganizing and Rejuvenating CNNs for Increasing Model Capacity Utilization
• Network Clustering for Multi-task Learning
• Towards Searching Efficient and Accurate Neural Network Architectures in Binary Classification Problems
• Evaluating Online and Offline Accuracy Traversal Algorithms for k-Complete Neural Network Architectures
• Disentangled Neural Architecture Search
• How Convolutional Neural Network Architecture Biases Learned Opponency and Colour Tuning
• Efficient Sampling for Predictor-Based Neural Architecture Search
• Efficient falsification approach for autonomous vehicle validation using a parameter optimisation technique based on reinforcement learning
• Reducing Inference Latency with Concurrent Architectures for Image Recognition
• Meta-learning Transferable Representations with a Single Target Domain
• AutoPruning for Deep Neural Network with Dynamic Channel Masking
• Deployment Optimization for Shared e-Mobility Systems with Multi-agent Deep Neural Search
• Improving the sample-efficiency of neural architecture search with reinforcement learning
• Conceptual Expansion Neural Architecture Search (CENAS)
• Multiple Myeloma Cancer Cell Instance Segmentation
• Rapid Model Architecture Adaption for Meta-Learning
• Automated Deep Abstractions for Stochastic Chemical Reaction Networks
• Neural Inheritance Relation Guided One-Shot Layer Assignment Search
• WAS-VTON: Warping Architecture Search for Virtual Try-on Network
• FLASH: Fast Neural Architecture Search with Hardware Optimization
• ADWPNAS: Architecture-Driven Weight Prediction for Neural Architecture Search
• Soft Layer Selection with Meta-Learning for Zero-Shot Cross-Lingual Transfer
• MONCAE: Multi-Objective Neuroevolution of Convolutional Autoencoders
• Encoder-Decoder Neural Architecture Optimization for Keyword Spotting
• Design of Artificial Intelligence Agents for Games using Deep Reinforcement Learning
• Mise en abyme with artificial intelligence: how to predict the accuracy of NN, applied to hyper-parameter tuning
• MLFriend: Interactive Prediction Task Recommendation for Event-Driven Time-Series Data
• Functional Neural Networks for Parametric Image Restoration Problems
• Learning Compatible Embeddings
• Communication-Efficient Separable Neural Network for Distributed Inference on Edge Devices
• An Embedded Deep Learning based Word Prediction
• Searching Learning Strategy with Reinforcement Learning for 3D Medical Image Segmentation
• Patterns versus Characters in Subword-aware Neural Language Modeling
• NeuralArTS: Structuring Neural Architecture Search with Type Theory
• Input-to-Output Gate to Improve RNN Language Models
• Channel Planting for Deep Neural Networks using Knowledge Distillation
• Multi-headed Neural Ensemble Search
• On the potential for open-endedness in neural networks
• Differentiable Architecture Pruning for Transfer Learning
• CompConv: A Compact Convolution Module for Efficient Feature Learning
• Mutually-aware Sub-Graphs Differentiable Architecture Search
• Poisoning the Search Space in Neural Architecture Search
• A Reinforcement Learning Approach for Sequential Spatial Transformer Networks
• Rethinking Layer-wise Feature Amounts in Convolutional Neural Network Architectures
• LV-BERT: Exploiting Layer Variety for BERT
• Neural Architecture Search with an Efficient Multiobjective Evolutionary Framework
• Transform-Based Feature Map Compression for CNN Inference
• A2: Extracting Cyclic Switchings from DOB-nets for Rejecting Excessive Disturbances
• Third ArchEdge Workshop: Exploring the Design Space of Efficient Deep Neural Networks
• Differentiable Neural Architecture Search with Morphism-based Transformable Backbone Architectures
• Architecture Agnostic Neural Networks
• BraidNet: procedural generation of neural networks for image classification problems using braid theory
• Fixed Priority Global Scheduling from a Deep Learning Perspective
• Evolving Neural Architecture Using One Shot Model
• Neural Architecture Search via Bregman Iterations
• Out-of-the-box channel pruned networks
• Recommending Courses in MOOCs for Jobs: An Auto Weak Supervision Approach
• DICE: Deep Significance Clustering for Outcome-Aware Stratification
• Estimating air quality co-benefits of energy transition using machine learning
• Unchain the Search Space with Hierarchical Differentiable Architecture Search
• Efficient Transfer Learning via Joint Adaptation of Network Architecture and Weight
• Understanding Cache Boundness of ML Operators on ARM Processors
• $S^2$-LBI: Stochastic Split Linearized Bregman Iterations for Parsimonious Deep Learning
• Inferring Mesoscale Models of Neural Computation
• Generalized Reinforcement Meta Learning for Few-Shot Optimization
• Photofeeler-D3: A Neural Network with Voter Modeling for Dating Photo Impression Prediction
• Enhanced Gradient for Differentiable Architecture Search
• Seeing Convolution Through the Eyes of Finite Transformation Semigroup Theory: An Abstract Algebraic Interpretation of Convolutional Neural Networks
• AmoebaContact and GDFold: a new pipeline for rapid prediction of protein structures
• Self-Constructing Neural Networks Through Random Mutation
• Learning, transferring, and recommending performance knowledge with Monte Carlo tree search and neural networks
• CP-NAS: Child-Parent Neural Architecture Search for Binary Neural Networks
• ARMIN: Towards a More Efficient and Light-weight Recurrent Memory Network
• Improving Generative Adversarial Networks with Local Coordinate Coding
• Knowledge Distillation By Sparse Representation Matching
• MTL2L: A Context Aware Neural Optimiser
• ENOS: Energy-Aware Network Operator Search for Hybrid Digital and Compute-in-Memory DNN Accelerators
• Multi-Objective DNN-based Precoder for MIMO Communications
• Neural Architecture Search for Image Super-Resolution Using Densely Constructed Search Space: DeCoNAS
• Making Differentiable Architecture Search less local
• Balancing Accuracy and Latency in Multipath Neural Networks
• Neural Architecture Search in operational context: a remote sensing case-study
• Edge-featured Graph Neural Architecture Search
• A Neural Architecture Search based Framework for Liquid State Machine Design
• Self-Supervised Neural Architecture Search for Imbalanced Datasets
• Max and Coincidence Neurons in Neural Networks
• Single-DARTS: Towards Stable Architecture Search
• Efficient Modelling Across Time of Human Actions and Interactions
• Multi-Faceted Hierarchical Multi-Task Learning for a Large Number of Tasks with Multi-dimensional Relations
• FOX-NAS: Fast, On-device and Explainable Neural Architecture Search
• Knowledge accumulating: The general pattern of learning
• Entropy Non-increasing Games for the Improvement of Dataflow Programming