This document introduces deep reinforcement learning and provides some examples of its applications. It begins with backgrounds on the history of deep learning and reinforcement learning. It then explains the concepts of reinforcement learning, deep learning, and deep reinforcement learning. Some example applications are controlling building sway, optimizing smart grids, and autonomous vehicles. The document also discusses using deep reinforcement learning for robot control and how understanding the principles can help in problem setting.
ゼロから始める深層強化学習(NLP2018講演資料)/ Introduction of Deep Reinforcement LearningPreferred Networks
Introduction of Deep Reinforcement Learning, which was presented at domestic NLP conference.
言語処理学会第24回年次大会(NLP2018) での講演資料です。
http://www.anlp.jp/nlp2018/#tutorial
This document discusses the relationship between control as inference, reinforcement learning, and active inference. It provides an overview of key concepts such as Markov decision processes (MDPs), partially observable MDPs (POMDPs), optimality variables, the evidence lower bound (ELBO), variational inference, and the free energy principle as applied to active inference. Control as inference frames reinforcement learning as probabilistic inference by defining a generative process and performing variational inference to find an optimal policy. Active inference uses the free energy principle and minimizes expected free energy to select actions that resolve uncertainty.
The document discusses distances between data and similarity measures in data analysis. It introduces the concept of distance between data as a quantitative measure of how different two data points are, with smaller distances indicating greater similarity. Distances are useful for tasks like clustering data, detecting anomalies, data recognition, and measuring approximation errors. The most common distance measure, Euclidean distance, is explained for vectors of any dimension using the concept of norm from geometry. Caution is advised when calculating distances between data with differing scales.
20220617_You_Only_Look_Once_Series.pdf
You Only Look Once: Unified, Real-Time Object Detection
https://www.cv-foundation.org/openaccess/content_cvpr_2016/html/Redmon_You_Only_Look_CVPR_2016_paper.html
YOLO9000: Better, Faster, Stronger
https://openaccess.thecvf.com/content_cvpr_2017/html/Redmon_YOLO9000_Better_Faster_CVPR_2017_paper.html
YOLOv3: An Incremental Improvement
https://arxiv.org/abs/1804.02767
YOLOv4: Optimal Speed and Accuracy of Object Detection
https://arxiv.org/abs/2004.10934
YOLOv5
https://github.com/ultralytics/yolov5
YOLOX: Exceeding YOLO Series in 2021
https://arxiv.org/abs/2107.08430
You Only Look One-Level Feature
https://openaccess.thecvf.com/content/CVPR2021/html/Chen_You_Only_Look_One-Level_Feature_CVPR_2021_paper.html
You Only Watch Once: A Unified CNN Architecture for Real-Time Spatiotemporal Action Localization
https://openaccess.thecvf.com/content/ICCV2021/html/Chen_Watch_Only_Once_An_End-to-End_Video_Action_Detection_Framework_ICCV_2021_paper.html
Managing Machine Learning workflows on Treasure DataAki Ariga
The document discusses managing machine learning workflows on Treasure Data. It describes Treasure Data's machine learning capabilities including its GUI interface, SQL queries integrated with workflows, and bundling of the Apache Hivemall library. It provides an example SQL query for training a supervised learning model and discusses how Treasure Workflow can be used to parameterize and parallelize ML workflows. Potential use cases for the new py> operator which allows Python scripts to be run on Treasure Data are also presented.
ゼロから始める深層強化学習(NLP2018講演資料)/ Introduction of Deep Reinforcement LearningPreferred Networks
Introduction of Deep Reinforcement Learning, which was presented at domestic NLP conference.
言語処理学会第24回年次大会(NLP2018) での講演資料です。
http://www.anlp.jp/nlp2018/#tutorial
This document discusses the relationship between control as inference, reinforcement learning, and active inference. It provides an overview of key concepts such as Markov decision processes (MDPs), partially observable MDPs (POMDPs), optimality variables, the evidence lower bound (ELBO), variational inference, and the free energy principle as applied to active inference. Control as inference frames reinforcement learning as probabilistic inference by defining a generative process and performing variational inference to find an optimal policy. Active inference uses the free energy principle and minimizes expected free energy to select actions that resolve uncertainty.
The document discusses distances between data and similarity measures in data analysis. It introduces the concept of distance between data as a quantitative measure of how different two data points are, with smaller distances indicating greater similarity. Distances are useful for tasks like clustering data, detecting anomalies, data recognition, and measuring approximation errors. The most common distance measure, Euclidean distance, is explained for vectors of any dimension using the concept of norm from geometry. Caution is advised when calculating distances between data with differing scales.
20220617_You_Only_Look_Once_Series.pdf
You Only Look Once: Unified, Real-Time Object Detection
https://www.cv-foundation.org/openaccess/content_cvpr_2016/html/Redmon_You_Only_Look_CVPR_2016_paper.html
YOLO9000: Better, Faster, Stronger
https://openaccess.thecvf.com/content_cvpr_2017/html/Redmon_YOLO9000_Better_Faster_CVPR_2017_paper.html
YOLOv3: An Incremental Improvement
https://arxiv.org/abs/1804.02767
YOLOv4: Optimal Speed and Accuracy of Object Detection
https://arxiv.org/abs/2004.10934
YOLOv5
https://github.com/ultralytics/yolov5
YOLOX: Exceeding YOLO Series in 2021
https://arxiv.org/abs/2107.08430
You Only Look One-Level Feature
https://openaccess.thecvf.com/content/CVPR2021/html/Chen_You_Only_Look_One-Level_Feature_CVPR_2021_paper.html
You Only Watch Once: A Unified CNN Architecture for Real-Time Spatiotemporal Action Localization
https://openaccess.thecvf.com/content/ICCV2021/html/Chen_Watch_Only_Once_An_End-to-End_Video_Action_Detection_Framework_ICCV_2021_paper.html
Managing Machine Learning workflows on Treasure DataAki Ariga
The document discusses managing machine learning workflows on Treasure Data. It describes Treasure Data's machine learning capabilities including its GUI interface, SQL queries integrated with workflows, and bundling of the Apache Hivemall library. It provides an example SQL query for training a supervised learning model and discusses how Treasure Workflow can be used to parameterize and parallelize ML workflows. Potential use cases for the new py> operator which allows Python scripts to be run on Treasure Data are also presented.
The document discusses machine learning projects and production. It begins with an introduction of Aki Ariga and their background. It then discusses 4 patterns for machine learning projects: 1) train batch/predict online via REST API, 2) train/predict batch via shared DB, 3) train/predict/serve continuously via streaming, and 4) train batch/predict on mobile apps. The document also covers machine learning operations (MLOps) including continuous integration/delivery, monitoring, feedback loops, and collaboration between researchers, developers and operations.
Why I started Machine Learning Casual Talks? #MLCTAki Ariga
This document discusses the start of the author's Machine Learning Casual Talks series. It aims to share practical know-how about machine learning applications that is often omitted from academic papers. Topics will include automated testing, key performance indicators, moving beyond just accuracy metrics, and the importance of development data. The author hopes to encourage sharing of real-world experience from their own work at Cookpad.
This document discusses an upcoming Ruby conference in Tokyo and provides related links. It mentions the Kawasaki Ruby user group and a Facebook group for parent engineers. A link is included for a timer tool called Rubyistimer that now has a gong feature, and there is a call to organize a Kanagawa Ruby conference.