Multi-Task Learning with Deep Neural Networks: A Survey

Abstract

Multi-task learning (MTL) is a subfield of machine learning in which multiple tasks are simultaneously learned by a shared model. Such approaches offer advantages like improved data efficiency, reduced overfitting through shared representations, and fast learning by leveraging auxiliary information. However, the simultaneous learning of multiple tasks presents new design and optimization challenges, and choosing which tasks should be learned jointly is in itself a non-trivial problem. In this survey, we give an overview of multi-task learning methods for deep neural networks, with the aim of summarizing both the well-established and most recent directions within the field. Our discussion is structured according to a partition of the existing deep MTL techniques into three groups: architectures, optimization methods, and task relationship learning. We also provide a summary of common multi-task benchmarks.

References in corpus (53)

• Sequence to Sequence Learning with Neural Networks
• Natural Language Processing (almost) from Scratch
• Neural Architecture Search with Reinforcement Learning
• Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
• Categorical Reparameterization with Gumbel-Softmax
• An Overview of Multi-Task Learning in Deep Neural Networks
• Estimating or Propagating Gradients Through Stochastic Neurons for Conditional Computation
• Microsoft COCO Captions: Data Collection and Evaluation Server
• The Lottery Ticket Hypothesis: Finding Sparse, Trainable Neural Networks
• Recurrent Neural Network for Text Classification with Multi-Task Learning
• End-To-End Memory Networks
• PathNet: Evolution Channels Gradient Descent in Super Neural Networks
• Solving Rubik's Cube with a Robot Hand
• A Survey on Multi-Task Learning
• IMPALA: Scalable Distributed Deep-RL with Importance Weighted Actor-Learner Architectures
• Multi-Task Learning Using Uncertainty to Weigh Losses for Scene Geometry and Semantics
• Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learning
• The Natural Language Decathlon: Multitask Learning as Question Answering
• Multi-Task Learning as Multi-Objective Optimization
• One Model To Learn Them All
• A Comprehensive Survey on Transfer Learning
• Multi-Task Deep Neural Networks for Natural Language Understanding
• Modular Multitask Reinforcement Learning with Policy Sketches
• Actor-Mimic: Deep Multitask and Transfer Reinforcement Learning
• Learning Multiple Tasks with Multilinear Relationship Networks
• Improving Multi-Task Deep Neural Networks via Knowledge Distillation for Natural Language Understanding
• AdaShare: Learning What To Share For Efficient Deep Multi-Task Learning
• Taking Advantage of Sparsity in Multi-Task Learning
• Gradient Surgery for Multi-Task Learning
• Learning and Transfer of Modulated Locomotor Controllers
• Adversarial Multi-task Learning for Text Classification
• Trace Norm Regularised Deep Multi-Task Learning
• Multi-Task Reinforcement Learning with Soft Modularization
• A Joint Many-Task Model: Growing a Neural Network for Multiple NLP Tasks
• Pareto Multi-Task Learning
• Routing Networks and the Challenges of Modular and Compositional Computation
• Routing Networks: Adaptive Selection of Non-linear Functions for Multi-Task Learning
• Modular Networks: Learning to Decompose Neural Computation
• Deep Multi-Task Learning with Shared Memory
• Adaptive Scheduling for Multi-Task Learning
• Modular meta-learning
• Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering
• Deep Model Transferability from Attribution Maps
• Learning Sparse Sharing Architectures for Multiple Tasks
• Exploring Shared Structures and Hierarchies for Multiple NLP Tasks
• OmniNet: A unified architecture for multi-modal multi-task learning
• Identifying beneficial task relations for multi-task learning in deep neural networks
• Feature Partitioning for Efficient Multi-Task Architectures
• Learning to Multi-Task by Active Sampling
• Meta Reinforcement Learning with Task Embedding and Shared Policy
• Multi-task Learning for Multi-modal Emotion Recognition and Sentiment Analysis
• Transfer Learning to Learn with Multitask Neural Model Search
• Learning Functions to Study the Benefit of Multitask Learning

Cited by in corpus (40)

• A Review on Deep Learning in UAV Remote Sensing
• Multimodal Co-learning: Challenges, Applications with Datasets, Recent Advances and Future Directions
• MEDIC: A Multi-Task Learning Dataset for Disaster Image Classification
• UAVs Meet LLMs: Overviews and Perspectives Toward Agentic Low-Altitude Mobility
• DeepMTS: Deep Multi-task Learning for Survival Prediction in Patients with Advanced Nasopharyngeal Carcinoma using Pretreatment PET/CT
• Multi-task fully convolutional network for tree species mapping in dense forests using small training hyperspectral data
• Debiasing Methods for Fairer Neural Models in Vision and Language Research: A Survey
• Recommender systems based on graph embedding techniques: A comprehensive review
• Depth Estimation and Image Restoration by Deep Learning from Defocused Images
• High-fidelity Prediction of Megapixel Longitudinal Phase-space Images of Electron Beams using Encoder-Decoder Neural Networks
• Graph Neural Networks for Surfactant Multi-Property Prediction
• Study of Subjective and Objective Quality Assessment of Mobile Cloud Gaming Videos
• Multitask Learning for Scalable and Dense Multilayer Bayesian Map Inference
• A store-and-forward cloud-based telemonitoring system for automatic assessing dysarthria evolution in neurological diseases from video-recording analysis
• INTERN: A New Learning Paradigm Towards General Vision
• Unified Question Generation with Continual Lifelong Learning
• Multi-head Sequence Tagging Model for Grammatical Error Correction
• Learning a Universal Template for Few-shot Dataset Generalization
• Aggregate or Not? Exploring Where to Privatize in DNN Based Federated Learning Under Different Non-IID Scenes
• Multi-task learning from fixed-wing UAV images for 2D/3D city modeling
• A voxel-level approach to brain age prediction: A method to assess regional brain aging
• Combiner and HyperCombiner Networks: Rules to Combine Multimodality MR Images for Prostate Cancer Localisation
• Spatio-Temporal Momentum: Jointly Learning Time-Series and Cross-Sectional Strategies
• STG-MTL: Scalable Task Grouping for Multi-Task Learning Using Data Map
• Modular Federated Learning
• Pedestrian Detection in Low-Light Conditions: A Comprehensive Survey
• ImageNet-21K Pretraining for the Masses
• SLAW: Scaled Loss Approximate Weighting for Efficient Multi-Task Learning
• T3S: Improving Multi-Task Reinforcement Learning with Task-Specific Feature Selector and Scheduler
• ClimateGAN: Raising Climate Change Awareness by Generating Images of Floods
• Tackling the Overestimation of Forest Carbon with Deep Learning and Aerial Imagery
• Self-Evolutionary Optimization for Pareto Front Learning
• Software/Hardware Co-design for Multi-modal Multi-task Learning in Autonomous Systems
• Meta-Learning Strategies through Value Maximization in Neural Networks
• Scalable Unidirectional Pareto Optimality for Multi-Task Learning with Constraints
• Provable Multi-Task Representation Learning by Two-Layer ReLU Neural Networks
• Accent Recognition with Hybrid Phonetic Features
• Fast Line Search for Multi-Task Learning
• Balancing Average and Worst-case Accuracy in Multitask Learning
• Attention-based multi-task learning for speech-enhancement and speaker-identification in multi-speaker dialogue scenario