|
Machine Learning Weekly Digest.
|
|
Welcome to this week of the Best of Machine Learning Digest. In this weekly newsletter, we
resurface some of the best resources in Machine Learning posted in the past week. This time,
we've gotten 61 submissions, including 3 papers.
|
Papers
This week, 3
Papers were posted on Best of ML. In
the
following, we're showing you the Top 3 posts of this week.
|
|
Null It Out: Guarding Protected Attributes
The ability to control for the kinds of information encoded in neural representation has a variety of use cases, especially in light of the challenge of interpreting these models. We present Iterative Null-space Projection (INLP), a novel method for removing information from neural representations. Our method is based on repeated training of linear classifiers that predict a certain property we aim to remove, followed by projection of the representations on their null-space.
|
|
Semantically Multi-modal Image Synthesis
In this paper, we focus on semantically multi-modal image synthesis (SMIS) task, namely, generating multi-modal images at the semantic level. Previous work seeks to use multiple class-specific generators, constraining its usage in datasets with a small number of classes. We instead propose a novel Group Decreasing Network (GroupDNet) that leverages group convolutions in the generator and progressively decreases the group numbers of the convolutions in the decoder.
|
|
ResNeSt: Split-Attention Networks
While image classification models have recently continued to advance, most downstream applications such as object detection and semantic segmentation still employ ResNet variants as the backbone network due to their simple and modular structure. We present a simple and modular Split-Attention block that enables attention across feature-map groups. By stacking these Split-Attention blocks ResNet-style, we obtain a new ResNet variant which we call ResNeSt.
|
Projects
This week, 24
Projects were posted on Best of ML. In
the
following, we're showing you the Top 2 posts of this week.
|
|
Minigrad: Autograd engine for Python written in Python C-API for speed, inspired by Karpathy's micrograd library
I like using my own libraries while learning stuff. I got excited when I saw Karpathy's small neural network library but it was a little bit slow for my taste. So I reimplemented it in Python C-API. It runs signifanctly faster. Of course, I would use PyTorch or Tensorflow for real projects but I learned how to implement an autograd engine and how to use Python C-API so it was worth it.
|
|
Convex Optimization: Algorithms and their Rate of Convergence
For anyone interested in learning more about convex optimization and it's theoretical guarantees below is a presentation on two major convex optim algorithms, Projected Gradient Descent and Mirror Descent.
|
Blogs
This week, 2
Blogs were posted on Best of ML. In
the
following, we're showing you the Top 2 posts of this week.
|
|
Logistic Regression
The idea behind logistic regression and the way its model works in classification problem and it get lowest cost value.
|
|
Gradient Descent Algorithm (part-1)
How Gradient Descent Algorithm works to get best value of "intercept-coeff." that satisfies the min. cost.
|
Blog Posts
This week, 31
Blog Posts were posted on Best of ML. In
the
following, we're showing you the Top 2 posts of this week.
|
|
PCA (Principle Component Analysis)
The basic idea of features reduction using principle component analysis
|
|
Understanding ACGANs with code [PyTorch]
ACGAN stands for Auxiliary Classifier Generative Adversarial Network. The network was developed by a group of researchers from google brain and was presented in the 34th international conference of machine learning held at Sydney, Australia. This article is a brief description of the research work enunciated in the paper and implementation of the same using PyTorch.
|
|
|
|
|
