Improve Supervised Models With Unlabeled Data

Показать описание

📈 You can leverage unlabelled data to improve supervised models with Self-Supervised Learning.

First, train your model to learn the generic pattern in the unlabelled data.

For instance, with images, you can hide random parts of the image and ask the model to reconstruct the missing piece.
Same for NLP with text.

Then, use the embedding learned in the first step and fine-tune the model on the data labels.

You can get better performance than using only the labeled data.

But it works better if the quantity of unlabeled data is huge to learn meaningful patterns.

Hopefully you liked this video 💚
🔥 Subscribe to Bitswired
👍🏽 Leave us a like/comment to support us

💬HASHTAGS:
#machinelearning #deeplearning #selfsupervisedlearning #data #datascience #ai #optimization #training #pytorch #tensorflow #python #programming #coding

Bitswired

Рекомендации по теме

Комментарии

Would you consider self-supervised learning for your next project?

bitswired

Improve Supervised Models With Unlabeled Data

Improve Supervised Models With Unlabeled Data

Supervised vs. Unsupervised Learning

Unit 7.7 | Using Unlabeled Data with Self-Supervised Learning | Part 1

Big Self-Supervised Models are Strong Semi-Supervised Learners (Paper Explained)

Supervised vs Unsupervised vs Reinforcement Learning | Machine Learning Tutorial | Simplilearn

MixMatch: A Holistic Approach to Semi-Supervised Learning

What is Semi-Supervised Learning?

Sukmin Yun: OpenCoS: Contrastive Semi-supervised Learning for Handling Open-set Unlabeled Data

6 Supervised Unsupervised Learning

Semi-Supervised Learning: the Case When Unlabeled Data is Equally Useful

Bhavana Dalvi: Semi Supervised Learning In Realistic Settings

Semi-Supervised Learning to Improve Detections - CVsports 2022 - #7

Labeled data vs Unlabeled Data

All Machine Learning algorithms explained in 17 min

Unlabeled Samples Generated by GAN Improve the Person Re-identification Baseline in vitro

Learning from Unlabeled Videos for Recognition, Prediction, and Control

Self Supervised Learning Predictive Maintenance | PhD Thesis #ai #phd #python #deeplearning

Self-Damaging Contrastive Learning Explained!

Leveraging Unlabeled Image Data With Self-Supervised Learning or Pseudo Labeling With Mateusz Opala

FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence

Self-training with Noisy Student improves ImageNet classification (Paper Explained)

Learning to Detect Important People in Unlabelled Images for Semi-Supervised Important People...

HyperTransformer: Model Generation for Supervised and Semi-Supervised Few-Shot Learning (w/ Author)

Machine Learning Series 5 : Semi- Supervised Learning Algorithms and Applications.