Training Data

Training data is crucial for machine learning model development, with current research focusing on improving data quality, efficiency, and mitigating biases. Active areas include generating synthetic data to address scarcity or privacy concerns, developing algorithms to optimize data selection and usage (e.g., self-paced learning, active learning), and mitigating issues like data contamination and imbalance through techniques such as data augmentation, selective parameter merging, and novel loss functions. The quality and characteristics of training data significantly impact model performance, generalization, and robustness, influencing various applications from natural language processing and image recognition to scientific computing and medical diagnosis.

Papers

July 20, 2022

A temporally and spatially local spike-based backpropagation algorithm to enable training in hardware
Anmol Biswas, Vivek Saraswat, Udayan Ganguly
Training Data Spiking Neural Network Spike Train Hardware Environment Spike Encoding Spike Based Backpropagation Spike Based Neuromorphic Spike Based Learning

July 15, 2022

PodcastMix: A dataset for separating music and speech in podcasts
Nicolás Schmidt, Jordi Pons, Marius Miron
Data Set Training Data Speech Analysis Source Separation Music Industry Background Music

July 14, 2022

Benign, Tempered, or Catastrophic: A Taxonomy of Overfitting
Neil Mallinar, James B. Simon, Amirhesam Abedsoltan, Parthe Pandit, Mikhail Belkin, Preetum Nakkiran
Training Data Comprehensive Taxonomy Neural Tangent Kernel Model Overfitting Benign Overfitting Benign Input Laplacian Kernel Exponential Measure

July 11, 2022

Horizontal Federated Learning and Secure Distributed Training for Recommendation System with Intel SGX
Siyuan Hui, Yuqiu Zhang, Albert Hu, Edmund Song
Deep Learning Training Data Recommendation System Secure Approach Deep Learning Recommendation Model Horizontal Federated Learning Intel Xeon

July 10, 2022

One-shot Neural Backdoor Erasing via Adversarial Weight Masking
Shuwen Chai, Jinghui Chen
Deep Neural Network Training Data Adversarial Mask Backdoor Removal Neural Backdoor

July 8, 2022

Ablation Study of How Run Time Assurance Impacts the Training and Performance of Reinforcement Learning Agents
Nathaniel Hamilton, Kyle Dunlap, Taylor T Johnson, Kerianne L Hobbs
Reinforcement Learning Training Data System Performance Safe Reinforcement Learning Reinforcement Learning Agent Policy Reinforcement Learning Safe Exploration Ablation Study

July 7, 2022

July 5, 2022

Cooperate or Compete: A New Perspective on Training of Generative Networks
Ch. Sobhan Babu, Ravindra Guravannavar, Arvind Hulgeri
Training Data GAN Model Long Term Cooperation Competition Platform New Perspective Generative Network Discriminator Guidance Simultaneous Move Game

July 4, 2022

Memory Efficient Patch-based Training for INR-based GANs
Namwoo Lee, Hyunsu Kim, Gayoung Lee, Sungjoo Yoo, Yunjey Choi
Training Data GAN Model Implicit Neural Representation Patch Based Memory Efficiency Efficient GAN

July 3, 2022

DiSCoMaT: Distantly Supervised Composition Extraction from Tables in Materials Science Articles
Tanishq Gupta, Mohd Zaki, Devanshi Khatsuriya, Kausik Hira, N. M. Anoop Krishnan, Mausam
Training Data NLP Task Table Semantics Material Science Compositional Data Table Representation

June 26, 2022

Class Impression for Data-free Incremental Learning
Sana Ayromlou, Purang Abolmaesumi, Teresa Tsang, Xiaoxiao Li
Training Data Incremental Learning Incremental Learning Framework Class Conditional Impression Reappearing Class Data

June 20, 2022

Measuring the Effect of Training Data on Deep Learning Predictions via Randomized Experiments
Jinkun Lin, Anqi Zhang, Mathias Lecuyer, Jinyang Li, Aurojit Panda, Siddhartha Sen
Deep Learning Model Training Data Mixed Effect Causal Inference Shapley Value Randomized Experiment LASSO Regression

Training Data

Papers

A temporally and spatially local spike-based backpropagation algorithm to enable training in hardware

PodcastMix: A dataset for separating music and speech in podcasts

Benign, Tempered, or Catastrophic: A Taxonomy of Overfitting

Wakeword Detection under Distribution Shifts

Normalized gradient flow optimization in the training of ReLU artificial neural networks

Developing a Component Comment Extractor from Product Reviews on E-Commerce Sites

Horizontal Federated Learning and Secure Distributed Training for Recommendation System with Intel SGX

One-shot Neural Backdoor Erasing via Adversarial Weight Masking

Ablation Study of How Run Time Assurance Impacts the Training and Performance of Reinforcement Learning Agents

Training Transformers Together

Neural Language Models are not Born Equal to Fit Brain Data, but Training Helps

FedHeN: Federated Learning in Heterogeneous Networks

Cooperate or Compete: A New Perspective on Training of Generative Networks

Memory Efficient Patch-based Training for INR-based GANs

DiSCoMaT: Distantly Supervised Composition Extraction from Tables in Materials Science Articles

Class Impression for Data-free Incremental Learning

EpiGRAF: Rethinking training of 3D GANs

Imitation Learning for Nonprehensile Manipulation through Self-Supervised Learning Considering Motion Speed

Propagation with Adaptive Mask then Training for Node Classification on Attributed Networks

Measuring the Effect of Training Data on Deep Learning Predictions via Randomized Experiments