陳葦庭,把機械業當飯店業拚,吉輔刀庫賣贏德國、日本,商業周刊,2021 年 03 月 29 日
3/30/2021
3/26/2021
Matrix Multiplication Inches Closer to Mythic Goal
Kevin Hartnett, Matrix Multiplication Inches Closer to Mythic Goal, Quanta Magazine, March 23, 2021.
“Exponent two” refers to the ideal speed — in terms of number of steps required — of performing one of the most fundamental operations in math: matrix multiplication. If exponent two is achievable, then it’s possible to carry out matrix multiplication as fast as physically possible. If it’s not, then we’re stuck in a world misfit to our dreams.
3/24/2021
跨越深度學習技術產品化的障礙
徐宏民,跨越深度學習技術產品化的障礙,電子時報,2021-03-24
一般而言,攔阻這些研究成為產品的障礙包括:運算速度/耗能、模型/演算法設計、場域穩定度(正確率)、訓練資料等。過去4年我們有機會在各式運算平台(雲端、Arm 、DSP、或是客製化的加速器上)實現各種智慧產品,希望這些經驗可以消除某些疑慮。
3/21/2021
國家的決斷
張國城,國家的決斷:給台灣人看的二戰後國際關係史,八旗文化,2019
《國家的決斷》作者張國城教授是知名的國防外交與軍事戰略專家。深諳國際關係現實主義理論的他,主張「國家」的能力、意圖與決斷是一切國際政治的根本。在本書中,他詳細分析從二戰結束後,從日本投降、聯合國與北約的建立、自由與共產陣營的分裂對峙、核武問世、鐵幕降臨、韓戰、越戰,一直到蘇聯解體、美國獨霸、第三波民主化與經濟全球化席捲全球,在這當中國家的興衰起伏與權力的消長更迭。而最重要的是,對於國家地位不明、被排除於國際組織之外的台灣來說,我們的安全與福祉如何被發生在其他國家的戰場與外交會議中的事件所決定?這是這本書試圖解答的。
「以台灣的人口、經濟規模和所處位置,卻沒有辦法參與絕大多數的國際組織,且幾乎無法發展正常的國與國關係,被世界上九成以上的國家否認,這不是正常狀況。」但缺少國家身分對台灣究竟帶來哪些傷害?我們對國際現實如何才能有清明的認識?政治領導人能否在需要捍衛國家利益時做出勇敢且前瞻的決斷?長期以來由於台灣遠離國際社會、對國際事務陌生而冷漠,我們常飽受缺乏「國際觀」之譏。《國家的決斷》從台灣人的視角出發,以台灣的利益為主體,屢屢剖析台灣如何在險惡的國際社會中走到今天,不僅彌補台灣人走向世界時的知識貧乏,更是思考台灣未來前途時必備的參考。
3/15/2021
從人機介面推論一家公司如何走向衰敗
6 年前買的冰箱,壓縮機一直發出巨大聲響,冷藏室不夠冷,於是上網填寫維修單。
網路上的表格,要求輸入室內電話。因為,只有行動電話,所以該欄位空白。送出後,要求一定要填寫輸入電話,只好再輸入假的室內電話 1234567,送出後,竟然說驗證碼錯誤,原來下方的驗證碼又重新改了,同時把原本輸入的所有資料刪除。如果各位有在高鐵上面訂票,欄位輸入錯誤,驗證碼是不會動的。無可奈何,只好再輸入一次。
3/13/2021
3/12/2021
大學申請面試流言終結運動
書審資料不做會比不上人家嗎?幹部證明及志工服務到底有沒有用?國立台北大學社會學系系主任陳婉琪,在臉書發起「申請面試流言終結運動」,邀請大學教授一起澄清社會大眾對大學申請常見的誤解。她表示,一旦想要跟風,你就失去了獨特的你自己。包括台大電機系教授葉丙成等上百名大學教師,都認同並分享此文。...
3/07/2021
周品均抓出團隊痛點讓美妝品牌從大虧到盈利
程倚華,別人不敢她敢!周品均上任唯品風尚CEO抓出團隊哪6大痛點,讓這家美妝品牌從大虧到盈利?,數位時代,2021.01.22
問題1:廣告策略
周品均觀察,起初,4大美妝品牌的廣告費佔了營收將近40%,ROAS(Return On Ad Spend,目標廣告支出回報率)卻只有0.8~1.1。
Learning Demand Curves in B2B Pricing
Huashuai Qu, Ilya O. Ryzhov, Michael C. Fu, Eric Bergerson Megan, and Kurka Ludek Kopacek, Learning Demand Curves in B2B Pricing: A New Framework and Case Study, Production and Operations Management, Volume 29, Issue 5, May 2020, Pages: 1287-1306.
In business-to-business (B2B) pricing, a seller seeks to maximize revenue obtained from high-volume transactions involving a wide variety of buyers, products, and other characteristics. Buyer response is highly uncertain, and the seller only observes whether buyers accept or reject the offered prices. These deals are also subject to high opportunity cost, since revenue is zero if the price is rejected. The seller must adapt to this uncertain environment and learn quickly from new deals as they take place. We propose a new framework for statistical and optimal learning in this problem, based on approximate Bayesian inference, which has the ability to measure and update the seller’s uncertainty about the demand curve based on new deals. In a case study, based on historical data, we show that our approach offers significant practical benefits.
3/06/2021
中鋼如何用AI煉成智慧鋼廠
翁芊儒,組織、人才和技術5年布局, 中鋼如何用AI煉成智慧鋼廠,iThome,2021-03-04
這一場變革,莫約從5年前開始推展。「我們展開數位轉型,是為了提升鋼鐵生產效率、降低成本、縮短交期,來提升產品競爭力。」中鋼技術部門代理副總經理鄭際昭,一句話點出轉型任務最重要的目的。...
3/02/2021
車用晶片缺貨
Debby Wu, Gabrielle Coppola, and Keith Naughton, A Year of Poor Planning Led to Carmakers’ Massive Chip Shortage, Bloomberg, 2021/1/19.
Near-sighted planning, supply-chain complexities and a tradition of keeping inventories low caused the semiconductor shortage that is now forcing carmakers to idle production lines and straining their relationship with chip manufacturers.
3/01/2021
Bearing Brings AI-Powered Operational Efficiencies to the Maritime Shipping Industry
Bearing.ai, Bearing Brings AI-Powered Operational Efficiencies to the Maritime Shipping Industry, February 16, 2021.
Bearing exits stealth mode and launches its AI-driven operations optimization platform, which provides a wide range of actionable insights to shipping companies, leading to improved efficiency, safety and reduced gas greenhouse emissions. Bearing’s platform is powered by highly accurate ship performance models built on real-world data, allowing it to predict fuel consumption, speed, and other performance factors much more accurately than existing solutions on the market. (e.g., Bearing’s typical prediction accuracy for fuel consumption is over 98% per voyage, compared to a typical existing accuracy of 80%).