投稿時間:2022-10-04 04:27:07 RSSフィード2022-10-04 04:00 分まとめ(31件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
AWS	AWS News Blog	AWS Week in Review – October 3, 2022	https://aws.amazon.com/blogs/aws/aws-week-in-review-october-3-2022/	AWS Week in Review October This post is part of our Week in Review series Check back each week for a quick roundup of interesting news and announcements from AWS A new week and a new month just started Curious which were the most significant AWS news from the previous seven days I got you covered with this post Last …	2022-10-03 18:32:35
AWS	AWS Database Blog	Stream data with Amazon DocumentDB and Amazon MSK using a Kafka connector	https://aws.amazon.com/blogs/database/stream-data-with-amazon-documentdb-and-amazon-msk-using-a-kafka-connector/	Stream data with Amazon DocumentDB and Amazon MSK using a Kafka connectorA common trend in modern application development and data processing is the use of Apache Kafka as a standard delivery mechanism for your data pipeline and fan out approach Amazon Managed Streaming for Apache Kafka Amazon MSK is a fully managed highly available and secure service that makes it simple for developers and DevOps managers to …	2022-10-03 18:29:50
AWS	AWS	AWS Config Conformance Packs Provide Scores To Help You Track Resource Compliance	https://www.youtube.com/watch?v=0OerXUxVG08	AWS Config Conformance Packs Provide Scores To Help You Track Resource ComplianceAWS Config now supports compliance scores as an enhancement to conformance packs With this video you ll understand how compliance scores offer a consistent measurement to track remediation progress perform comparisons across different sets of requirements and see the impact a specific change or deployment has on your compliance posture Learn more at Subscribe More AWS videos More AWS events videos ABOUT AWSAmazon Web Services AWS is the world s most comprehensive and broadly adopted cloud platform offering over fully featured services from data centers globally Millions of customers ーincluding the fastest growing startups largest enterprises and leading government agencies ーare using AWS to lower costs become more agile and innovate faster AWSConfig AWS AmazonWebServices CloudComputing	2022-10-03 18:01:16
AWS	AWS - Webinar Channel	Analytics in 15: Observability and Operational Analytics with AWS	https://www.youtube.com/watch?v=hosEb1HDBEo	Analytics in Observability and Operational Analytics with AWSData collection and monitoring in one place enables full stack end to end observability of your entire infrastructure The Amazon OpenSearch Service includes an observability interface and log monitoring features that provide developers and dev ops engineers with the visibility and insights they need to diagnose performance and availability issues faster and reduce application downtime Learning Objectives Objective Correlate trace data with log events and navigate multiple interfaces Objective Visualize data driven events using Piped Processing Language Objective Learn about the new Observability interface and OpenSearch Notebooks to interact and collaborate to develop rich reports To learn more about the services featured in this talk please visit	2022-10-03 18:17:19
AWS	AWS - Webinar Channel	Analytics in 15: No-code/Low-code ETL and Data Integration with AWS	https://www.youtube.com/watch?v=S-iAjjxEbbU	Analytics in No code Low code ETL and Data Integration with AWSETL the easy way AWS Glue AWS Glue DataBrew and AWS Glue Studio offer developers and data engineers an easy no code and low code experience to working with data transformations across complex data pipelines Learn more about the serverless ETL capabilities of AWS Glue Learning Objectives Objective Build and run transforms quickly in a serverless fashion across data pipelines Objective See how data analysts can clean and enrich data for downstream use cases such as machine learning Objective Learn how to integrate with a diverse set of data sources to create an end to end data pipeline with very little coding effort To learn more about the services featured in this talk please visit	2022-10-03 18:11:37
海外TECH	DEV Community	Meme Monday 🦊	https://dev.to/ben/meme-monday-33ej	Meme Monday Welcome to another Meme Monday post Today s cover image comes from last week s thread DEV is an inclusive space Humor in poor taste will be downvoted by mods	2022-10-03 18:43:21
海外TECH	DEV Community	Hacktoberfest 2022	https://dev.to/dtslvr/hacktoberfest-2022-441l	Hacktoberfest We are very excited to join Hacktoberfest for the first time with Ghostfolio and meet new and ambitious open source contributors Hacktoberfest is a month long celebration of open source projects their maintainers and the entire community of contributors Each October open source maintainers from all over the world give extra attention to new contributors while guiding them through their first pull requests on GitHub About GhostfolioGhostfolio is a modern web application to manage your personal finance The software presents the current assets in real time and supports the decision making of future investments Whether rebalancing the asset classes stocks ETFs cryptocurrencies etc of your portfolio or financing an apartment Ghostfolio offers solid data driven decision support Ghostfolio is written in TypeScript and organized as an Nx workspace The backend is based on NestJS using PostgreSQL as a database together with Prisma and Redis for caching The frontend is built with Angular How to contribute Every contribution matters This can be implementing new features fixing bugs refactoring the code improving the documentation adding more unit tests or translating into another language Are you not yet familiar with our code base That is not a problem We have applied the label hacktoberfest to a few issues and ideas that are well suited for newcomers The official Hacktoberfest website provides some valuable resources for beginners to start contributing in open source Get supportIf you have further questions or ideas please join our growing Slack community or get in touch on Twitter ghostfolio We look forward to hearing from you Thomas from Ghostfolio	2022-10-03 18:23:50
海外TECH	DEV Community	Deep Learning Library From Scratch 7: Implementing RNN layers	https://dev.to/ashwinscode/deep-learning-library-from-scratch-7-implementing-rnn-layers-1ea2	Deep Learning Library From Scratch Implementing RNN layersHi guys Welcome back to part of this series of where we build our own deep learning library Last article went through the implementation of the automatic differentiation module We will see how easy this module makes adding new layer types by implementing RNN layers in this part of the series The code for this series library can be found at this repo ashwins code Zen Deep Learning Library Deep Learning library written in Python Contains code for my blog series on building a deep learning library Zen Deep Learning LibraryA deep learning library written in Python Contains the code for my blog series where we build this library from scratchmnist py contains an example for a MNIST digit recogniser using the libraryrnn py contains an example of a Recurrent Neural Network which learns to fit to the graph sin x cos x View on GitHub What are RNNs So far in our library we have only encountered simple feed forward neural networks These networks work well in quite a few cases but when our inputs introduce the concept of time feed forward networks begin to struggle since they contain no mechanism to encapsulate contextual data Several pieces of data like language and stock price data just to name a few are determined from historical data points For example in terms of generating language the words in a sentence are determined by what words have already been generated In terms of predicting stock prices the previous prices can be used to determine whether the stock price rises falls Our feed forward networks simply do not have the mechanisms to handle such types of data so how can we model such data RNNs were designed for this specific problem RNN stands for recurrent neural network As the name suggests RNNs rely on a recurrence mechanism to capture contextual data How do they work RNNs utilise a hidden state to encapsulate sequential data What the hidden state exactly is will be explained a bit later RNNs take in a sequence as input They can output both a new sequence and the final hidden state from the input x is the input sequencey is the outputted sequenceh is the final hidden stateU W V are all linear functionsHow does this exactly happen I like to think of RNNs as a unit that is applied to each time step of the input sequence A time step refers to the item in a sequence at a specific point in time xtx txt​ refers to the item of the input sequence xxx at time step ttt yty tyt​ refers to the item of the output sequence yyy at time step ttt hth tht​ refers to the hidden state after iterating through t t t time stepsAs the diagram shows RNN units produce an output sequence time step and a hidden state after being fed the input sequence time step and the previous hidden state using the linear functions UUU WWW VVV UUU WWW VVV take the form f x Wx Bf x Wx Bf x Wx Bwhere WWW is some adjustable weight matrix and xxx and BBB are vectors BBB is an adjustable vector The hidden state is a vector that encapsulates the meaning of the sequence It essentially carries the information of all the time steps it has seen so far yty tyt​ and hth tht​ are caculated as follows u U xt w W ht ht tanh u w yt V ht u U x t newlinew W h t newlineh t tanh u w newliney t V h t newlineu U xt​ w W ht ​ ht​ tanh u w yt​ V ht​ Hopefully you can see that if these calculations are applied to each time step of the input sequence an output sequence and a final hidden state is produced This final hidden state is a vector that contains all the information about the sequence the RNN layer was given The output sequence could be given to another RNN layer to reveal more complex patterns within the data During training the weights and biases used within the UUU WWW and VVV functions are adjusted so that the output sequence and hidden state better represent the inputted sequence Code ImplementationWe shall add our RNN layer class to nn py import autodiff as adimport numpy as npimport loss import optim class RNN Layer def init self units hidden dim return sequences False self units units self hidden dim hidden dim self return sequences return sequences self U None self W None self V None def one forward self x x np expand dims x axis state np zeros x shape self hidden dim y for time step in x mul u self U time step mul w self W state state Tanh mul u mul w if self return sequences y append self V state if not self return sequences state value state value squeeze return state return y def call self x if self U is None self U Linear self hidden dim self W Linear self hidden dim self V Linear self units if not self return sequences states for seq in x state self one forward seq states append state s ad stack states return s sequences for seq in x out seq self one forward seq sequences append out seq return sequences def update self optim self U update optim self W update optim if self return sequences self V update optim Let s break this down into its separate methods def init self units hidden dim return sequences False self units units self hidden dim hidden dim self return sequences return sequences self U None self W None self V NoneThe class constructor takes in three parameters units the size of each time step in the outputted sequence when return sequences is truehidden dim the size of the hidden statereturn sequences if true the RNN layer will return a newly calculated sequence the same length as the input sequence If false it returns the final hidden stae self U Noneself W Noneself V NoneU W V are initialised as None but will represent the different weight matrices shown earlier after the layer s first forward pass The following methods have been commented to be better understood def call self x This method takes in a batch of sequences and returns the final hidden states after iterating through each sequence if return sequences is False otherwise it returns the output sequences if self U is None intialise U W V if this is first forward pass Since we know U W V are all linear functions with trainable parameters we can simply assign them to instances of our already existing Linear class self U Linear self hidden dim self W Linear self hidden dim self V Linear self units if not self return sequences states go through each sequence for seq in x apply the one forward method to sequence state self one forward seq append the final hidden state states append state use stack method to convert these list of tensors into a single tensor so that its derivative can be calculated s ad stack states return hidden states return s sequences go through each sequence for seq in x apply the one forward method to sequence out seq self one forward seq append the output sequence sequences append out seq return output sequences return sequencesdef one forward self x This method takes in a list representing a single sequence x np expand dims x axis making list numpy array with dimensions so that its shape can be calculated for the following line state np zeros x shape self hidden dim hidden state intitialised as a matrix of s y this array will store the output sequence if return sequences is True iterate through each time step in the sequence for time step in x perform next state calculation mul u self U time step mul w self W state state Tanh mul u mul w if self return sequences calculate the output sequence time step if return sequences is True y append self V state if not self return sequences return hidden state if return sequence is False state value state value squeeze return state return output sequence is return sequences is True return ydef update self optim self U update optim self W update optim if self return sequences self V update optim This update method was seen in the linear layer too This method is called during backpropagation to adjust the weights of the layers You may have noticed a new function from our autodiff module stack Here is the code for the stack function autodiff pydef stack tensor list tensor values tensor value for tensor in tensor list s np stack tensor values var Tensor s var dependencies tensor list for tensor in tensor list var grads append np ones tensor value shape return varThis joins a list of tensors together into a single tensor This makes it possible for a group of separately computed tensors needed to be operated on at once while still recording the operation onto the computation graph Notice how we did not need to write any backpropagation code for this class Our autodiff module provides that layer of abstraction for us making implementing new layer types much easier Building an RNN model to test it all out To see if this all is working well we are going to build an RNN model to extrapolate trigonometric curves The RNN will take in a sequence of y values from consecutive points from the curve It s job is to predict the next y value of the next point on the curve The RNN will extrapolate the curve by repeatedly feeding the most recent points on the curveImports import numpy as npimport nnimport optimimport lossfrom matplotlib import pyplot as pltGenerate the first points of a sin wavex axis np array list range Tseq np sin i for i in range Prepping dataset Inputs will contain a window of points from the curve The output is the y value of the next point on the curve x y for i in range len seq new seq i for i in seq i i x append new seq y append seq i Building modelmodel nn Model nn RNN hidden dim return sequences True nn RNN hidden dim units is since it is irrelevant return sequence is False by default nn Linear nn Linear Train model and predict The predictions are plotted to a graph and saved as a png file model train np array x np array y epochs optimizer optim RMSProp loss fn loss MSE batch size preds for i in x preds append model np expand dims np array i axis value plt plot x axis seq plt plot x axis preds plt savefig graph png Running the code will results in something similar to the following EPOCH ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ lt it s LOSS EPOCH ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ lt it s LOSS EPOCH ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ lt it s LOSS EPOCH ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ lt it s LOSS EPOCH ███████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████ lt it s LOSS ResultsThe results for a few expressions have been shown below Blue represents the true valuesOrange represents the model s predictionssin x sin x cos x I am happy with these results Thanks for reading Thanks for reading through this blog post We will continue to add new layer types in the next few posts of this series so be ready for those	2022-10-03 18:02:05
Apple	AppleInsider - Frontpage News	Will Smith's 'Emancipation' gets unexpected Apple TV+ release date	https://appleinsider.com/articles/22/10/03/will-smiths-emancipation-gets-unexpected-apple-tv-release-date?utm_medium=rss	Will Smith x s x Emancipation x gets unexpected Apple TV release dateDelayed by a lawsuit and Will Smith s infamous Oscar slap Apple TV has announced Emancipation will be in theaters and streaming in December Emancipation has not had the easiest production with COVID halting filming and producers moving shooting away from its Georgia location over the state s voting laws Then came a sexual harassment lawsuit and Will Smith causing controversy by hitting Chris Rock during the Academy Awards After that Oscars slap Variety said that it was the unspoken truth that the film was being pushed back into to distance it from the news However Apple s has seen its hopes for back to back Oscar wins ーfollowing CODA in ーfading after critical flops such as Luck and The Greatest Beer Run Ever Read more	2022-10-03 18:07:18
海外科学	NYT > Science	Unearthing Everyday Life at an Ancient Site in Greece	https://www.nytimes.com/2022/10/03/travel/iklaina-archaeology-greece.html	general	2022-10-03 18:20:06
海外TECH	WIRED	Success on Twitch No Longer Comes on Twitch	https://www.wired.com/story/twitch-streamers-leaving-tiktok-youtube/	twitch	2022-10-03 18:34:32
海外TECH	WIRED	How to Shop Like a Pro During Amazon's Prime Early Access Sale	https://www.wired.com/story/amazon-prime-day-2022-tips-to-find-best-deals/	deals	2022-10-03 18:16:00
ニュース	@日本経済新聞 電子版	省エネ暖房、欧州に照準　パナソニックは生産能力6倍に https://t.co/OnV5bKFnLz	https://twitter.com/nikkei/statuses/1576995917932396545	生産能力	2022-10-03 18:01:40
ニュース	@日本経済新聞 電子版	［社説］首相は強い指導力で政策実行を着実に https://t.co/xNAGZmZonH	https://twitter.com/nikkei/statuses/1576995917018042369	首相	2022-10-03 18:01:40
ニュース	@日本経済新聞 電子版	重い日本の学費・新車販売増加・「かっぱ寿司」社長辞任 https://t.co/S2rt7RVTHU	https://twitter.com/nikkei/statuses/1576995916099166211	販売	2022-10-03 18:01:40
ニュース	@日本経済新聞 電子版	ローカル線、存続なら｢6割値上げ｣の未来 https://t.co/YaibMCvar2	https://twitter.com/nikkei/statuses/1576995915000295425	存続	2022-10-03 18:01:39
ニュース	@日本経済新聞 電子版	マック全店で紙ストロー導入　脱プラ、年900トン削減 https://t.co/IQ8GFj4f7S	https://twitter.com/nikkei/statuses/1576995914010746880	導入	2022-10-03 18:01:39
ニュース	@日本経済新聞 電子版	日経、デジタル3媒体を11月24日創刊　車･GX･先端技術 https://t.co/S3KHcVyke1	https://twitter.com/nikkei/statuses/1576995913024708608	先端技術	2022-10-03 18:01:39
ニュース	@日本経済新聞 電子版	政策へ勝負の3カ月　岸田政権1年、かすむ「黄金の3年」 https://t.co/Ia2Cm0a1Vv	https://twitter.com/nikkei/statuses/1576995912018055168	黄金	2022-10-03 18:01:39
ニュース	@日本経済新聞 電子版	ノーベル賞ペーボ氏　古代人DNAから現代人のルーツ探る https://t.co/hjVWwd4Mmr	https://twitter.com/nikkei/statuses/1576995910998929408	現代人	2022-10-03 18:01:39
ニュース	BBC News - Home	Gold coins hidden in 7th Century found in wall	https://www.bbc.co.uk/news/world-middle-east-63122180?at_medium=RSS&at_campaign=KARANGA	muslim	2022-10-03 18:48:01
ビジネス	ダイヤモンド・オンライン - 新着記事	稲盛和夫『働き方』1章（2）：ど真剣に働く――「人生を好転させる」法 - 稲盛和夫『働き方』1章を読む	https://diamond.jp/articles/-/310239	稲盛和夫	2022-10-04 03:50:00
ビジネス	ダイヤモンド・オンライン - 新着記事	現金でリターンを生む3つの方法 - WSJ PickUp	https://diamond.jp/articles/-/310699	wsjpickup	2022-10-04 03:45:00
ビジネス	ダイヤモンド・オンライン - 新着記事	米消費支出にのしかかる金利上昇 - WSJ PickUp	https://diamond.jp/articles/-/310700	wsjpickup	2022-10-04 03:40:00
ビジネス	ダイヤモンド・オンライン - 新着記事	貧困層が最も資産増加、米国「下からの回復」 - WSJ PickUp	https://diamond.jp/articles/-/310701	wsjpickup	2022-10-04 03:35:00
ビジネス	ダイヤモンド・オンライン - 新着記事	売上高7600億円企業の創業者が語る「危機を乗り越え成長する会社」の特徴とは - 経営の本質 すべては人に始まり人に終わる	https://diamond.jp/articles/-/309956	危機的状況	2022-10-04 03:30:00
ビジネス	ダイヤモンド・オンライン - 新着記事	ツイッター炎上の意外なメリット、監視社会は本当に悪なのか？ - 定番読書	https://diamond.jp/articles/-/309090	監視社会	2022-10-04 03:25:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「エリトリアってどんな国？」2分で学ぶ国際社会 - 読むだけで世界地図が頭に入る本	https://diamond.jp/articles/-/308707		2022-10-04 03:20:00
ビジネス	ダイヤモンド・オンライン - 新着記事	史上「人心操作」で悪用されてきたヤバい心理手法とは？ - 自分を変える方法	https://diamond.jp/articles/-/310071	行動科学	2022-10-04 03:15:00
ビジネス	ダイヤモンド・オンライン - 新着記事	【どちらが印象に残りやすい？】脳の習性を利用する！傷つきすぎた心にすぐに効く「自分を大切にする簡単な練習」 - あなたはもう、自分のために生きていい	https://diamond.jp/articles/-/310380	twitter	2022-10-04 03:10:00
ビジネス	ダイヤモンド・オンライン - 新着記事	「老害にならない人」の特徴とは？PDCAよりも今の時代に合った思考法 - アジャイル仕事術	https://diamond.jp/articles/-/310684		2022-10-04 03:05:00