Machine Learning

Recurrent Neural Networks were great in mid-2017. They were able to do everything a successful sequence model was supposed to do, albeit with some drawbacks. Then transformers (Attention Is All You Need) came along and soon enough every state-of-the-art model in NLP was a transformer. Recently with OpenAI’s ImageGPT and DALL-E, Transformers have been increasingly applied as an … [Read more...] about A Comprehensive Guide to Transformers (Part 1: The Encoder)

To address these problems we propose Elo-based Predictive Power (EPP) ranking method.Elo measureThe Elo ranking system is used for calculating the relative skill levels of players in games, such as chess or soccer but Elo is also popular in MOBA (Multiplayer online battle arena) games.The difference in Elo scores of two players is a predictor of their match result. Elo is … [Read more...] about Elo Ranks for Visual Task Adaptation Benchmark

Next Steps:This article is part of a mini-series that helps readers set up everything they need to start learning about artificial intelligence, machine learning, deep learning, and or data science. It includes articles that contain instructions with copy and paste code and screenshots to help readers get the outcome as soon as possible. It also includes articles that contain … [Read more...] about How to Install the Python Environment for AI and Machine Learning on Windows 10

Note: This is just a basic high-level explanation, it might not cover the topics in detail.Imagine you are in a car on top of a hill, your objective is to go down (reach the plains, aka the global minima). So what do you do?You use your 360 degree camera and see which path is the steepest — why? because you want to climb down as fast as possible (your GF/BF just called). So how … [Read more...] about Gradient descent & its DNN relatives — A short story

文章标题：How we use AutoML, Multi-task learning and Multi-tower models for Pinterest Ads解读人：爱晒太阳的小白猫简介：本文来自于pinterest广告组，主要讲解在工作中是如何使用auto ml技术、multi-task learning和multi-tower模型。既有宏观介绍，又覆盖了一些模型训练的小技巧，比如在实战中如何calibrate multi-task model的预测结果。阅读难易程度：简单Pinterest主要在特征工程和建模两方面使用了auto ml的技术。特征工程Pinterest使用AutoML技术来做很多feature engineering（特征工程）的工作。第一步是把特征分类：continuous：可以直接用的连续浮点数one … [Read more...] about 技术博客阅读笔记：揭秘Pinterest广告组的深度学习推荐模型实践

You can learn some of the top platforms at home and for FREE!Congratulations upcoming grads! You’ve sat through countless hours of classes, written your sorting algorithms from scratch, analyzed your lab data in Excel, and you’re about to receive that piece of paper stating that maybe you know a thing or two about your STEM fields.Regardless of your field, I don’t need to tell … [Read more...] about Data Analytics Resources for Students and Recent Grads

Necessity of antenna miniaturization in wireless communicationIn almost all areas of electrical engineering the research interest has been shifted toward miniaturization. Electromagnetism and antennas in particular, are no exception. With the rapid development in wireless systems all over the world, miniaturized antennas have become an active area of research for many antenna … [Read more...] about Deep learning for miniaturized slot antenna design using metamaterials inspired techniques

Se tem algo que está na moda atualmente é Machine Learning (em português, aprendizado de máquina). As máquinas roubarão seu emprego, saber mexer com isso garantirá seu futuro, tudo em pouco tempo será regido por ML ou AI (inteligência artificial)… Enfim, o que não falta são afirmações como estas vindas de vários tipos de pessoas, que sabem ou não sobre o assunto (como todo … [Read more...] about 3 coisas importantes de se fazer ao estudar Machine Learning

ここでは少し題材が不適切かと思ったのですが、COVIDの陽性者予測を実際にしてみます。データは厚生労働省に日次で全国の陽性者数が公開されているオープンデータがありますから、これを使ってみましょう。データはシンプルで、日付, 陽性者数のみが入っています。インポートする時はDATETIME型でインポートすると後工程が楽でしょう。データは2020年の1月1日から入っています。次にモデリングをします。Auto … [Read more...] about BigQuery MLのARIMAモデルを使ってCOVID陽性者数の時系列予測をする

How do neural networks behave when the hidden layers are very large? A new research direction in machine learning studies network behavior in the limit that the size of each hidden layer increases to infinity. Surprisingly, in this limit the behavior of the network dramatically simplifies and we can understand what the network is doing theoretically. Why does this happen and … [Read more...] about A Crash Course on Wide Neural Nets and the Neural Tangent Kernel