coral-cnn
PyTorchZeroToAll
coral-cnn | PyTorchZeroToAll | |
---|---|---|
4 | 1 | |
330 | 3,825 | |
2.1% | - | |
0.0 | 0.0 | |
about 3 years ago | almost 2 years ago | |
Python | Python | |
MIT License | - |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
coral-cnn
-
[D] Why is Ordinal Regression so overlooked?
The most recent and usable DL attempt I have found is the CORAL/CORN frameworks (keras, pytorch) which have just a few stars, and that's it.
-
[D] can regression models be used for ranking?
To your question, there are specific types of models called ordinal regression / ordinal classification models that do not assume a metric distance between values. E.g., if you have "20/hr, $15/hr, $0/hr" these models don't assume that the distance between 0 and 15 is 3x the distance between 20 and 15. It just assumes 20 > 15 > 0. We worked on this a bit in the context of neural networks: https://www.sciencedirect.com/science/article/pii/S016786552030413X , https://raschka-research-group.github.io/coral_pytorch/
-
[D] Modeling class errors
If you are interested, I recently worked on a simple ordinal regression approach for neural networks here: https://www.sciencedirect.com/science/article/pii/S016786552030413X
-
[R] [D] What machine learning methods can be used for ordinal regression?
Just took a quick look at that paper, it sounds like a good approach. If you are interested, we recently developed an ordinal regression approach with implementation in PyTorch (https://github.com/Raschka-research-group/coral-cnn). Someone also recently ported it to Keras: https://github.com/ck37/coral-ordinal. I haven't read the paper you mentioned in detail, but it seems our method is similar except that we add the probabilities that are >0.5 and that we have theoretical guarantees. rank consistency.
PyTorchZeroToAll
-
[D] I'm trying to do more stuff in pure Tensorflow. Is there an in-depth book that explain constructing recurrent, convolutional, graph etc layers in it?
I'm doing this rn, but with PyTorch. I look for notebooks/scripts (https://github.com/hunkim/PyTorchZeroToAll, primarily) on github read and copy them, along with the deeplearning book. Surprisingly pretty much everything just clicks now, it's my third attempt reading the text though. I don't think any book serves the purpose well, except for knowing the well established conventions of the field.
What are some alternatives?
coral-ordinal - Tensorflow Keras implementation of ordinal regression using consistent rank logits (CORAL) by Cao et al. (2019)
python-systemd-tutorial - A tutorial for writing a systemd service in Python
datatap-python - Focus on Algorithm Design, Not on Data Wrangling
python-minecraft-clone - Source code for each episode of my Minecraft clone in Python YouTube tutorial series.
contrastive-unpaired-translation - Contrastive unpaired image-to-image translation, faster and lighter training than cyclegan (ECCV 2020, in PyTorch)
vision_transformer - Discover how to build vision transformer from scratch with this comprehensive tutorial. Follow our step-by-step guide to create your own vision transformer.
ludwig - Low-code framework for building custom LLMs, neural networks, and other AI models
stanford-tensorflow-tutorials - This repository contains code examples for the Stanford's course: TensorFlow for Deep Learning Research.
horovod - Distributed training framework for TensorFlow, Keras, PyTorch, and Apache MXNet. [Moved to: https://github.com/horovod/horovod]
Hands-On-Deep-Learning-Algorithms-with-Python - Hands-On Deep Learning Algorithms with Python, By Packt
Calculadora - Calculadora simples escrita em Python. Realiza soma, subtração, multiplicação, divisão, exponenciação e raiz.
QualityScaler - QualityScaler - image/video deeplearning upscaling for any GPU