作者staff23 (筆名蘇溫永(星星送行者))
看板L_TalkandCha
標題Fw: [公告] 懶叫水桶 ^Q^
時間Fri Jun 29 13:04:00 2018
※ [本文轉錄自 staff23 信箱]
作者: RS5566 (懶叫好吃^Q^) 看板: Gossiping
標題: [公告] 懶叫水桶 ^Q^
時間: Tue Feb 27 20:37:56 2018
02/27/2018 20:36:47 RS5566 暫停
tcfd817038 發言，期限為 180 天
理由: 濫用爆卦
02/27/2018 20:36:59 RS5566 暫停
victor77 發言，期限為 90 天
理由: 政問
02/27/2018 20:37:12 RS5566 暫停
justlive 發言，期限為 90 天
理由: 字數未滿
02/27/2018 20:37:32 RS5566 暫停
stuj9019 發言，期限為 180 天
理由: 手動置底
作者 tcfd817038 (~喵星人~☆) 看板 Gossiping
標題 [爆卦] 地震
時間 Mon Feb 26 23:32:38 2018
───────────────────────────────────────
有感地震!!!
一直在搖
https://i.imgur.com/JzTk1ab.jpg
各位鍵盤邊緣人還好嗎?
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 36.239.124.87
※ 文章網址: https://www.ptt.cc/bbs/Gossiping/M.1519659163.A.A4D.html
作者 victor77 (bird) 看板 Gossiping
標題 [問卦] 228為什麼KMT有錯
時間 Sun Feb 25 21:19:12 2018
───────────────────────────────────────
台灣人的祖國日本打了敗仗
把台灣讓給中國
那島上的台灣人就是戰俘了
中國人就算殺死台灣人
強姦台灣妹子
那也是日本戰敗得付出的代價
戰爭的本質就是如此
現在戰敗國回來檢討戰勝國
說要轉型正義
那怎麼不先要求美國為兩顆原子彈道歉？
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 101.8.231.64
※ 文章網址: https://www.ptt.cc/bbs/Gossiping/M.1519564755.A.EB9.html
作者 justlive (不要跟豬吵架) 看板 Gossiping
標題 Re: [ＦＢ] 柯文哲【明天開始上班!】
時間 Wed Feb 21 17:53:22 2018
───────────────────────────────────────
9 common mistakes executives make with data
Mariya Yao@thinkmariya February 19, 2018 2:10 PM
https://venturebeat.com/2018/02/19/9-common-mistakes-executives-make-with-data
Data is a human invention. Humans define the phenomenon they want to measure,
design systems to collect data about it, clean and pre-process it before
analysis, and finally, choose how to interpret the results. Even with the
same dataset, two people can form vastly different conclusions. This is
because data alone is not “ground truth” — observable, provable, and
objective data that reflects reality. If researchers infer data from other
information, rely on subjective judgment, do not collect data in a rigorous
and careful manner, or use sources that are of questionable authenticity,
then the data they produce it is not ground truth.
How you choose to conceptualize a phenomenon, determine what to measure, and
decide how to take measurements will affect the data that you collect. Your
ability to solve a problem with artificial intelligence depends heavily on
how you frame your problem and whether you can establish ground truth without
ambiguity. We use ground truth as a benchmark to assess the performance of
algorithms. If your gold standard is wrong, then your results will not only
be wrong but also potentially harmful to your business.
Unless you were directly involved with defining and monitoring your original
data collection goals, instruments, and strategy, you are likely missing
critical knowledge that may result in incorrect processing, interpretation,
and use of that data.
What people call “data” can actually be things like carefully curated
measurements selected purely to support an agenda; haphazard collections of
random information with no correspondence to reality; or information that
looks reasonable but resulted from unconsciously biased collection efforts.
Here’s a crash course on nine common statistical errors that every executive
should be familiar with.
1. Undefined goals
Failing to pin down the reason for collecting data means that you’ll miss
the opportunity to articulate assumptions and to determine what to collect.
The result is that you’ll likely collect the wrong data or incomplete data.
A common trend in big data is for enterprises to gather heaps of information
without any understanding of why they need it and how they want to use it.
Gathering huge but messy volumes of data will only impede your future
analytics, since you’ll have to wade through much more junk to find what you
actually want.
2. Definition error
Let’s say you want to know how much your customers spent on your services
last quarter. Seems like an easy task, right? Unfortunately, even a simple
goal like this will require defining a number of assumptions before you can
get the information that you want.
First, how are you defining “customer”? Depending on your goals, you might
not want to lump everyone into one bucket. You may want to segment customers
by their purchasing behaviors in order to adjust your marketing efforts or
product features accordingly. If that’s the case, then you’ll need to be
sure that you’re including useful information about the customer, such as
demographic information or spending history.
There are also tactical considerations, such as how you define quarters. Will
you use fiscal quarters or calendar quarters? Many organizations’ fiscal
years do not correspond with calendar years. Fiscal years also differ
internationally, with Australia’s fiscal year starting on July 1 and India’
s fiscal year starting on April 1. You will also need to develop a strategy
to account for returns or exchanges. What if a customer bought your product
in one quarter but returned it in another? What if they filed a quality
complaint against you and received a refund? Do you net these in the last
quarter or this one?
As you can see, definitions are not so simple. You will need to discuss your
expectations and set appropriate parameters in order to collect the
information you actually want.
3. Capture error
Once you’ve identified the type of data that you wish to collect, you’ll
need to design a mechanism to capture it. Mistakes here can result in
capturing incorrect or accidentally biased data. For example, if you want to
test whether product A is more compelling than product B, but you always
display product A first on your website, then users may not see or purchase
product B as frequently, leading you to the wrong conclusion.
4. Measurement error
Measurement errors occur when the software or hardware you use to capture
data goes awry, either failing to capture usable data or producing spurious
data. For example, you might lose information about user behavior on your
mobile app if the user experiences connectivity issues and the usage logs are
not synchronized with your servers. Similarly, if you are using hardware
sensors like a microphone, your audio recordings may capture background noise
or interference from other electrical signals.
5. Processing error
As you can see from our simple attempt to calculate customer sales earlier,
many errors can occur even before you look at your data. Many enterprises own
data that is decades old, where the original team capable of explaining their
data decisions is long gone. Many of their assumptions and issues are likely
not documented and will be up to you to deduce, which can be a daunting task.
You and your team may make assumptions that differ from the original ones
made during data collection and achieve wildly different results. Common
errors include missing a particular filter that researchers may have used on
the data, using different accounting standards, and simply making
methodological mistakes.
6. Coverage error
Coverage error describes what happens with survey data when there is
insufficient opportunity for all targeted respondents to participate. For
example, if you are collecting data on the elderly but only offer a website
survey, then you’ll probably miss out on many respondents.
In the case of digital products, your marketing teams may be interested in
projecting how all mobile smartphone users might behave with a prospective
product. However, if you only offer an iOS app but not an Android app, the
iOS user data will give you limited insight into how Android users may behave.
7. Sampling error
Sampling errors occur when you analyze data from a smaller sample that is not
representative of your target population. This is unavoidable when data only
exists for some groups within a population. The conclusions that you draw
from the unrepresentative sample will probably not apply to the whole.
A classic example of a sampling would be to ask only your friends or peers
for opinions about your company’s products, then assume the user population
will feel similarly.
8. Inference error
Statistical or machine learning models make inference errors when they make
incorrect predictions from the available ground truth. False negatives and
false positives are the two types of inference errors that can occur. False
positives occur when you incorrectly predict that an item belongs in a
category when it does not. False negatives occur when an item is in a
category, but you predict that it is not.
Assuming you have a clean record of ground truth, calculating inference
errors will help you assess the performance of your machine learning models.
However, the reality is that many real-world datasets are noisy and may be
mislabeled, which means you may not have clarity on the exact inference
errors your AI system makes.
9. Unknown error
Reality can be elusive, and you cannot always establish ground truth with
ease. In many cases, such as with digital products, you can capture tons of
data about what a user did on your platform but not their motivation for
those actions. You may know that a user clicked on an advertisement, but you
don’t know how annoyed they were with it.
In addition to many known types of errors, there are unknowns about the
universe that leave a gap between your representation of reality, in the form
of data, and reality itself.
Executives without a data science or machine learning background often make
these nine major errors, but many more subtle issues can also thwart the
performance of AI technologies you build that make predictions from data.
Mariya Yao is the CTO of Metamaven, an applied AI firm building custom
automation solutions for marketing and sales, and the coauthor of Applied
Artificial Intelligence, a book for business leaders.
This story originally appeared on Www.metamaven.com. Copyright 2018
※ 引述《uus (亞典波羅)》之銘言：
: https://goo.gl/FY7ABq
: 世大運實際售票數字 驗證柯網紅行銷無效
: 文／林靖堂
: 世大運前，台北市政府以行銷世大運之名，找來蔡阿嘎等數名網紅、YouTuber與市
: 長柯文哲合作拍片，引發網紅行銷的討論，以及柯市長是否是在藉此行銷自己。然
: 而，世大運結束了，柯文哲當時的網紅行銷是捧紅了世大運還是捧紅自己？或許，
: 就讓世大運的官方銷售數字說話吧。
: 今年7月26日，世大運前夕柯文哲在臉書上釋出多支與網紅合作行銷的影片，包括蔡
: 阿嘎、阿滴英文、上班不要看、走路痛、星期天、HowHow等Youtuber，共同宣傳世
: 大運。
: 依市府提供之資料，北市府用了100萬，製作了這9部影片，但以最夯的蔡阿嘎影片
: 為例，連結到購票網站的點閱數是3787次（至7月28日止）；7月30日的累計觀看次
: 數為203萬餘，轉換率估計低於0.5%；若影片的效果隨時間遞減，以4800次點閱為估
: 計基礎，每次點閱的成本為20元。
: 民進黨市議員李慶鋒就質疑，每支影片花不到10萬元的確便宜，youtuber很賣力，
: 柯市長很配合，的確大部分人都覺得很好笑，看來笑果十足，但卻效果有限，怎麼
: 可能會這樣？
: https://cdn-images-1.medium.com/max/600/0*rb7P2GuPrj7FUKwg.jpg
: （圖：世大運開閉幕典禮售票趨勢統計）
: 然而，根據體育署8月23日所提供，世大運官網自6月13日至8月19日開幕典禮當天，
: 實際的開閉幕典禮售票數字。售票趨勢顯示，若網紅影片約在7月26日前後釋出，那
: 段時間的售票數字，並無太大變化，僅閉幕典禮的座位，在8月1日有小幅成長，當
: 天售出約1000票，但，那已經是網紅影片釋出約一周以後的事。
: 進一步分析，該網紅議題，是在7月25、26兩天開始在網路與傳統主流媒體間發酵。
: 依上表，可以看得出，就8月19日的開幕典禮而言，7月25日至8月19日間毫無起色。
: 不過，根據體育署資料，開幕式共可售出1萬2333席座位（含3500元、2000元與800
: 元），早於7月17日時，已售出1萬1956席，等於是在網紅議題發酵前，開幕式即已
: 近乎完售。
: https://cdn-images-1.medium.com/max/800/0*jCx0Vkipna3OQV8F.jpg
: （圖：開閉幕典禮售票趨勢，至7/17）
: 然而，在閉幕式總售票數1萬2780席中，7月17日時，仍僅售出2成，共2629張，到了
: 8月9日則來到56%共7227張。這段時間，是在8月1日單日售出超過1000張與8月7日單
: 日售出超過4000張。最後，在8月23日時，因世大運選手表現太好，讓閉幕式1萬2777
: 張票完售。
: https://cdn-images-1.medium.com/max/800/0*aOk74pxqDu6e52dt.jpg
: （圖：開閉幕與競賽銷售狀況 統計時間8/9）
: 根據上述開閉幕式售票數的的相關資料，似乎佐證，網紅影片雖然熱門，但並未達
: 成讓世大運售票率增加的行銷效果，因為從官網實際上獲得的行銷數據，並未有相
: 關正向反應。
: https://cdn-images-1.medium.com/max/800/0*bUpqKNTlmCl-9AJP.jpg
: (圖:7/17前 22項競賽門票銷售狀況統計)
: 而若我們再看22項競賽項目的細部銷售情況。如上，7月17日網紅行銷影片尚未釋出
: 前，22項競技賽的總售票數字，僅售出12萬3926張票（含套票），約僅總售票數的
: 一成四。
: https://cdn-images-1.medium.com/max/800/0*vaK4E36El_jorp0u.jpg
: (圖:7/31前 22項競賽門票銷售狀況統計)
: 到了8月9日，也就是網紅行銷兩周後，22項競技賽的票務總銷售，才售出23萬0281張
: 票（含套票），約莫為總售票數三成，這期間約增加10萬6355張票，僅約為總售票
: 數的一成二。
: https://cdn-images-1.medium.com/max/800/0*S-iQqG6DO1K6J8_h.jpg
: (圖:8/23前 22項競賽門票銷售狀況統計)
: 直到8月23日，賽事漸熱，售票才開始衝破7成（58萬0001張），8月27日，世大運總
: 售票數已賣出超過67萬5000張票，總售票比例高達81.6%。
: 回頭檢視，網紅行銷的時間點在7月25日，話題效果或可延燒至8月1日，但若7月17日
: 時，賽事僅售出12萬3926張票，約僅總票數的一成四，到了8月9日，才售出23萬0281張
: 票，等於是網紅行銷期間，才約增加一成二的售票數。實際上，網紅效益並未顯現。
: 必須說，對照實際售票的數字趨勢攤開來看，無論是在開閉幕的售票趨勢，以及22項
: 競賽的售票數字，皆顯示，網紅行銷世大運的效應其實不大。
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 140.116.253.71
※ 文章網址: https://www.ptt.cc/bbs/Gossiping/M.1519206805.A.78A.html
作者 stuj9019 (MLGPRO) 看板 Gossiping
標題 [問卦] 為啥高中英文沒免修門檻
時間 Mon Feb 26 09:01:12 2018
───────────────────────────────────────
吶吶吶 餓死抬頭
又是魯魯的高中生我啦
我姐成大醬料 多益945就能夠不用修英文課了
啊魯魯的高中生我
https://i.imgur.com/hOnJpMT.jpg
幹為啥我還要修英文啊 廢廢的南部高中>醬料？
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 1.200.44.86
※ 文章網址: https://www.ptt.cc/bbs/Gossiping/M.1519606874.A.9FC.html
作者 stuj9019 (MLGPRO) 看板 Gossiping
標題 [問卦] 為啥高中英文沒免修門檻
時間 Mon Feb 26 08:54:39 2018
───────────────────────────────────────
吶吶吶 餓死抬頭
又是魯魯的高中生我啦
我姐成大醬料 多益945就能夠不用修英文課了
啊魯魯的高中生我
https://i.imgur.com/hOnJpMT.jpg
幹為啥我還要修英文啊 廢廢的南部高中>醬料？
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 1.200.44.86
※ 文章網址: https://www.ptt.cc/bbs/Gossiping/M.1519606481.A.12C.html
--
※各位請注意!!!! 已有非法懶叫掠奪集團，開始行動
標題 [新聞] 裝熟惹厭 男遭烤肉醬刷LP
承辦警員說：「被害人生殖器差點就被當『香腸』烤！有夠離譜。」 2012 10/30
推 RS5566:各位要小心 已有集團開始掠奪懶覺了11/12 00:01
推 johnny:幹我以前都以為RS是亂講的 想不到真有這種集團11/12 00:14
→ ben54302:RS大PO了 這麼多呼籲懶覺文....真乃聖人也!!11/12 02:25
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 114.46.21.191
※ 文章網址: https://www.ptt.cc/bbs/Gossiping/M.1519735081.A.F2B.html
→ WolfTeacher: ^Q^ 02/27 20:38
→ wwvvkai: 發懶叫了 02/27 20:38
推 a81255316: 錢 02/27 20:38
→ leopika: 豪ㄘ 02/27 20:38
推 loveyoucheng: 發錢啊 02/27 20:38
推 pionaero: 發錢啊幹 02/27 20:38
推 tomwu1993: 導覽較^Q^ 02/27 20:38
噓 rriver: money 02/27 20:38
→ d125383957: 錢 02/27 20:38
推 zxasqw0246: 發前LA 02/27 20:38
推 Kitakami: ^Q^ 02/27 20:38
→ RS5566: 發樓下五位 之前四差屎贊助的 02/27 20:38
噓 ms0400281: 發錢 02/27 20:39
→ balcony5566: $$ 02/27 20:39
推 gvvg5566: 發懶叫喇^Q^ 02/27 20:39
推 love0504: 五樓吃六樓懶叫 02/27 20:39
推 a81255316: 錢 02/27 20:39
推 pipi8696044: 發錢 02/27 20:39
推 Crazy7oker: 好給我 02/27 20:39
推 B05504021: ya 02/27 20:39
→ B05504021: 幹沒搶到 02/27 20:40
→ Forthelife: 吃吃 02/27 20:40
→ Crazy7oker: 不!!!!!懶叫王我也要錢QQ 02/27 20:40
→ balcony5566: 收到ㄌYA 下次一定再投你 02/27 20:40
推 Ironstick: 發懶覺 02/27 20:40
02/27/2018 20:40:03 支出 $100 ($552 => $452) 轉帳給 Ikos2239 (稅後 $90)
02/27/2018 20:40:04 支出 $100 ($452 => $352) 轉帳給 ms0400281 (稅後 $90)
02/27/2018 20:40:05 支出 $100 ($352 => $252) 轉帳給 balcony5566 (稅後 $90)
02/27/2018 20:40:06 支出 $100 ($252 => $152) 轉帳給 gvvg5566 (稅後 $90)
02/27/2018 20:40:07 支出 $100 ($152 => $52) 轉帳給 love0504 (稅後 $90)
※ 編輯: RS5566 (114.46.21.191), 02/27/2018 20:41:01
→ ng860310: 發錢^Q^ 02/27 20:40
推 yonggan: 懶叫 02/27 20:41
推 EGOiST40: 發懶叫 02/27 20:41
推 a91031309: 發錢 02/27 20:41
→ johnwu: 我也要 02/27 20:41
推 gvvg5566: 收到喇~下次投你^Q^ 02/27 20:41
→ Benbenyale: QQQ 02/27 20:42
推 WeAntiTVBS: 那些鬧版的終身水桶也不足惜啦！！ 02/27 20:42
推 pipi8696044: 沒拿到 QQ 02/27 20:43
推 yzli: ^Q^ 02/27 20:44
噓 nomoneverdie: 發錢啦 02/27 20:46
推 starbucks16: 錢 02/27 20:47
推 poopoo86213: ㄑㄧㄢ 02/27 20:47
推 misu2718: 這標題 02/27 20:47
推 kenlin0105: 錢 02/27 20:48
噓 zxcvbbb: 滾啦幹 02/27 20:48
推 yu800910: 發錢 02/27 20:48
推 GodPenis: $$$$$ 02/27 20:49
推 HowToBasic: 乾 那長串是三小 02/27 20:49
→ BlackBass: 錢 02/27 20:49
推 justastupid: ^Q^ 02/27 20:51
推 lagidog: 發懶覺 02/27 21:28
推 ZhiLou: 高中生愛秀成績 結果被桶ㄏㄏ 02/27 22:06
→ lunawolke: 懶覺好吃^Q^ 02/27 22:13
→ e86117934: 懶叫啦！^Q^ 02/27 22:27
推 jjba310: 香～發懶覺^Q^ 02/28 03:07
※ 發信站: 批踢踢實業坊(ptt.cc)
※ 轉錄者: staff23 (1.160.79.161), 06/29/2018 13:04:00
