DIM: Learning Deep Representations by Mutual Information Estimation and Maximization
Author(s): Sherwin Chen Originally published on Towards AI. Encoder Network Source: Pixabay This is our second article of the series about mutual information. In the previous articles, we have seen how to maximizes the mutual information between two variables via the MINE …
DIAYN: Diversity Is All You Need
Author(s): Sherwin Chen Originally published on Towards AI. Top highlight Diving Into DIAYN U+007C Towards AI An Unsupervised Information-Based Method to Learn Diverse Skills Different skills learned by DIAYN without any extrinsic reward signal. Source: https://sites.google.com/view/diayn Introduction We discuss an information-based reinforcement …
Several Ways for Machine Learning Model Serving (Model as a Service)
Author(s): Edward Ma Originally published on Towards AI. Using Model as a Service (MaaS) on Cloud Platforms Top highlight Photo by Edward Ma on Unsplash No matter how well you build a model, no one knows it if you cannot ship model. …
@Bayes’ Theorem For Bae
Author(s): Michael Knight Originally published on Towards AI. Intro to Probability and Stats U+007C Towards AI Introduce someone to probability theory and statistics without scaring them off Source Bayes’ Theorem is something that confuses and frustrates many, but is not as awful …
How to Train MAML(Model-Agnostic Meta-Learning)
Author(s): Sherwin Chen Originally published on Towards AI. An elaborate explanation for MAML and more Top highlight Source: Pixabay Model-Agnostic Meta-Learning(MAML) has been growing more and more popular in the field of meta-learning since it’s first introduced by Finn et al. in …
EMI: Exploration with Mutual Information
Author(s): Sherwin Chen Originally published on Towards AI. A novel exploration method based on representation learning Source: Photo by Andrew Neel on Unsplash Reinforcement learning could be hard when the reward signal is sparse. In these scenarios, exploration strategy becomes essentially important: …
New Model for Word Embeddings which are Resilient to Misspellings (MOE)
Author(s): Edward Ma Originally published on Towards AI. Photo by Edward Ma on Unsplash Traditional word embeddings are good at solving lots of natural language processing (NLP) downstream problems such as documentation classification and named-entity recognition (NER). However, one of the drawbacks …
Artificial Intelligence Without the Utopian Promise-land and Dystopian Armageddon
Author(s): Davor Petreski Originally published on Towards AI. The Future of AI U+007C Towards AI Before you start reading, think of 3 possible scenarios for the future of Artificial Intelligence (AI). If I asked you to think of 3 possible scenarios for …
Stop Calling Automation AI… and the Natural Progression of Intelligent Machines
Author(s): Junis Alico Originally published on Towards AI. Defining Automation and AI U+007C Towards AI Photo credit: Image by Gerd Altmann In the last few years, I’ve noticed a lot of C-level executives use AI in their keynote speeches and television appearances. …
PEARL: Probabilistic Embeddings for Actor-Critic RL
Author(s): Sherwin Chen Originally published on Towards AI. A sample-efficient meta reinforcement learning method Top highlight Source: Unsplash Meta reinforcement learning could be particularly challenging because the agent has to not only adapt to the new incoming data but also find an …
Exploring Deep Dream using Tensorflow 2.0
Author(s): Abhinav Prakash Originally published on Towards AI. Diving Into Deep Dream using Tensorflow U+007C Towards AI Deep Dream Using Tensorflow My image which generated by Deep Dream. Whenever any person hears about Deep Learning or Neural Network the things which first …
How does Data Noising Help to Improve your NLP Model?
Author(s): Edward Ma Originally published on Towards AI. The objective of injecting data noising Photo by Edward Ma on Unsplash Introducing data noising to neural network aims at improving model generalization and performance. Xie et al. proposes several ways to generate more …
StyleGAN Generated Face Classification with ResNexts
Author(s): Luka Chkhetiani Originally published on Towards AI. Using ResNexts with StyleGAN U+007C Towards AI https://generated.photos A week or two back a team released a dataset of 100K images of generated faces, based on StyleGAN [Karras et al. and NVIDIA] (see paper). …
AI & Physics, Part 1 — Dark Matter Matters Too
Author(s): Ryan Lynch Originally published on Towards AI. Diving Into AI & Physics U+007C Towards AI Physics. Our way of understanding the universe around us. We use equations and formulae to try and break down the fundamental truths. To figure out the …
Review: DUNet — Deformable U-Net for Retinal Vessels Segmentation (Biomedical Image Segmentation)
Author(s): Sik-Ho Tsang Originally published on Towards AI. Technical Review of DUNet U+007C Towards AI U-Net+DCN, Outperforms U-Net & DCN In this story, DUNet, by Tianjin University, Linkoping University, and, is briefly reviewed. DUNet, Deformable U-Net: exploits the retinal vessels’ local features …
Recent Posts
