 |
 |
 |
 |
 |
 |
||
 |
 |
 |
 |
 |
| 5 | 1/1 | 返回列表 |
| 查看: 3563 | 回復(fù): 13 | |||
| 【懸賞金幣】回答本帖問(wèn)題，作者shangu111將贈(zèng)送您 500 個(gè)金幣 | |||
| 當(dāng)前只顯示滿足指定條件的回帖，點(diǎn)擊這里查看本話題的所有回帖 | |||
shangu111新蟲(chóng) (小有名氣)
|
[求助]
TKDE 兩個(gè)major revision 一個(gè)reject，最后意見(jiàn)為reject，想申述，請(qǐng)大神看看可能性 已有2人參與
|
||
|
去年投了IEEE Transactions on Knowledge and Data Engineering，一審花了一年，review 1 給的reject， 我覺(jué)得絕大部分意見(jiàn)都是錯(cuò)的，有些我專(zhuān)門(mén)用了一章節(jié)介紹，有些我根本沒(méi)做過(guò)那些假設(shè)， 有些是很正常的常識(shí)，所以很氣憤，感覺(jué)完全不尊重別人的成果。我接下來(lái)貼出review 1 的意見(jiàn)和我的簡(jiǎn)單的rebuttal， 想要大家?guī)臀铱纯瓷晔龅目赡苄浴＿�有想問(wèn)問(wèn)有沒(méi)有人向TKDE申述過(guò)。 The issues of this paper are discussed as below: 1. This paper claims that the trading behavior is a function of data cost. But they just take 1(accept), 0(otherwise) into consideration, which do not show any practical correlation with data cost. This contribution is weak and limited. Rebuttal: trading behavior 指的是賣(mài)(0)或者不賣(mài)(1), 沒(méi)有其他可能，trading behavior 都可以formulate成data cost 的一個(gè)function了，他們之間還沒(méi)有correlation么？ 2.Compared with the previous works, the authors not only protect the data’ but also protect the trading behaviors’ privacy under LDP framework. However, since the trading behaviors only have two kinds of results 1(accept), 0(otherwise), the real trading behaviors can be obtained by the collectors after trades. Rebuttal: 從來(lái)沒(méi)有reveal real trading behaviours, 一直reveal的,sellers執(zhí)行的都是perturbed trading decision。 3.The assumption 1 does not always hold since the data may be in a high dimensional space. And PCA cannot ensure any loss in information when projecting the high dimensional data into a lower dimensional space. Rebuttal：PCA 不就是這個(gè)用處么？保持90%或以上的variance （這就是保障沒(méi)有l(wèi)oss in information） 同時(shí)減少 data dimension 4.In order to obtain the unbiased gradient for the data collector, the authors show us that the data owners should provide two auxiliary parameters, which is calculated by the biased gradient, with the collector. Firstly, what is physical meaning of these two parameters? Why the data owners provides these two kinds of information to help collector guess their real data cost? Rebuttal：因?yàn)槟阋猟ata owners 要賣(mài)他的data啊，要提供auxiliary parameters是你賣(mài)data的前提要求啊，何況提供了auxiliary parameters你的data 也被LDP保護(hù) 5.Subsection 4.1 states that the data owners perturb the data by using the perturbed gradient. But why is directly perturbing the raw data inapplicable? The authors should attempt to use their designed algorithm to learn real data cost from the perturbed data instead of redesigning a data perturbation method fitting the pricing rule to help learn the real data cost. Rebuttal: 為什么不直接用已經(jīng)在文章里面單獨(dú)列了個(gè)section說(shuō)了，一般的learning algorithm over perturbed data都會(huì)對(duì)這個(gè)perturbed data做一些假設(shè)，比如符合什么分布，怎么perturbed的，我這篇文章就假設(shè)perturbed data是從我這個(gè)perturbed scheme 生產(chǎn)出來(lái)的不行么？ 6.In Algorithm 2, there is an assumption that the price follows uniform distribution U[0,2]. There is no reference and explanation to support this assumption. Rebuttal:根本沒(méi)這個(gè)假設(shè)�。。� In Eq(1), the pricing rule is published with the help of two parameters ϵ,δ. Is it reasonable for the data collector to get the value of ϵ,δ ? Rebuttal：ϵ,δ是data collector設(shè)置的，為啥他不知道？ In experiments, Fig.6 is so confusing. There is only one computation time for one trading in practice. Why the authors the CDF of computation time for one trading? Rebuttal：因?yàn)檎麄�(gè)learning algorithm的gradients 都是randomized，而且計(jì)算gradients還用了external library，我做1000 trails，然后畫(huà)cdf不是很正常么？ |
回復(fù)此樓» 猜你喜歡
275求調(diào)劑
已經(jīng)有9人回復(fù)
-
083000學(xué)碩274求調(diào)劑
已經(jīng)有5人回復(fù)
-
【求調(diào)劑】085601材料工程專(zhuān)碩 | 總分272 |
已經(jīng)有5人回復(fù)
-
275求調(diào)劑
已經(jīng)有13人回復(fù)
-
0856，材料與化工321分求調(diào)劑
已經(jīng)有8人回復(fù)
-
085701環(huán)境工程，267求調(diào)劑
已經(jīng)有12人回復(fù)
-
320分，材料與化工專(zhuān)業(yè)，求調(diào)劑
已經(jīng)有7人回復(fù)
-
085701環(huán)境工程求調(diào)劑
已經(jīng)有8人回復(fù)
-
295材料工程專(zhuān)碩求調(diào)劑
已經(jīng)有5人回復(fù)
-
265求調(diào)劑
已經(jīng)有8人回復(fù)
肥耳朵1895
新蟲(chóng) (著名寫(xiě)手)
- 應(yīng)助: 0 (幼兒園)
- 金幣: 4511.1
- 散金: 400
- 沙發(fā): 4
- 帖子: 2164
- 在線: 382小時(shí)
- 蟲(chóng)號(hào): 26941881
- 注冊(cè): 2021-08-04
- 專(zhuān)業(yè): 電力系統(tǒng)
|
你首先要平靜情緒，作者要做的不是跟他吵架，而是解釋清楚。審稿人不可能看的太仔細(xì)，作者需要反思自己的寫(xiě)作，可能并沒(méi)有能讓讀者輕松get到貢獻(xiàn)。你目前這個(gè)反駁，我估計(jì)編輯都不一定能滿意。首先要花一兩句表明文章的貢獻(xiàn)，之后再解釋細(xì)節(jié)。 發(fā)自小木蟲(chóng)IOS客戶端 |
4樓2021-10-30 12:45:58
2樓2021-10-30 12:39:09
肥耳朵1895
新蟲(chóng) (著名寫(xiě)手)
- 應(yīng)助: 0 (幼兒園)
- 金幣: 4511.1
- 散金: 400
- 沙發(fā): 4
- 帖子: 2164
- 在線: 382小時(shí)
- 蟲(chóng)號(hào): 26941881
- 注冊(cè): 2021-08-04
- 專(zhuān)業(yè): 電力系統(tǒng)
3樓2021-10-30 12:39:13
5樓2021-10-30 13:07:13
| 最具人氣熱帖推薦 [查看全部] | 作者 | 回/看 | 最后發(fā)表 | |
|---|---|---|---|---|
|
[考研] 求調(diào)劑 +4 | 零八# 2026-03-27 | 4/200 |
|
|---|---|---|---|---|
|
[考研] 考研調(diào)劑 +4 | Sanmu-124 2026-03-26 | 4/200 |
|
|
|
[考研] 求調(diào)劑推薦 材料 304 +14 | 荷包蛋hyj 2026-03-26 | 14/700 |
|
|
|
[考研] 0703化學(xué) +6 | 妮妮ninicgb 2026-03-27 | 6/300 |
|
|
|
[考研] 322求調(diào)劑 +4 | 我真的很想學(xué)習(xí) 2026-03-23 | 4/200 |
|
|
|
[考研] 359求調(diào)劑 +4 | 王了個(gè)楠 2026-03-25 | 4/200 |
|
|
|
[考研] 324求調(diào)劑 +8 | hanamiko 2026-03-26 | 10/500 |
|
|
|
[考研] 一志愿河工大 081700 276求調(diào)劑 +4 | 地球繞著太陽(yáng)轉(zhuǎn) 2026-03-23 | 4/200 |
|
|
|
[考研] 一志愿天津大學(xué)339材料與化工求調(diào)劑 +3 | 江往賣(mài)魚(yú) 2026-03-26 | 3/150 |
|
|
|
[考研] 309求調(diào)劑 +4 | gajsj 2026-03-25 | 5/250 |
|
|
|
[考研] 一志愿哈工大，085400，320，求調(diào)劑 +4 | gdlf9999 2026-03-24 | 4/200 |
|
|
|
[考研] 299求調(diào)劑 +4 | 15188958825 2026-03-25 | 4/200 |
|
|
|
[考研] 332求調(diào)劑 +6 | 032500 2026-03-25 | 6/300 |
|
|
|
[考研] 考研調(diào)劑 +6 | 來(lái)好運(yùn)來(lái)來(lái)來(lái) 2026-03-21 | 7/350 |
|
|
|
[考研] 機(jī)械學(xué)碩總分317求調(diào)劑！！�。� +4 | Acaciad 2026-03-25 | 4/200 |
|
|
|
[考研] 284求調(diào)劑 +15 | Zhao anqi 2026-03-22 | 15/750 |
|
|
|
[考研] 0703化學(xué)求調(diào)劑 +6 | 奶油草莓. 2026-03-22 | 7/350 |
|
|
|
[考研] 一志愿國(guó)科過(guò)程所081700，274求調(diào)劑 +3 | 三水研0水立方 2026-03-23 | 3/150 |
|
|
|
[考研] 328求調(diào)劑 +4 | LHHL66 2026-03-23 | 4/200 |
|
|
|
[考研] 求調(diào)劑 +5 | Zhangbod 2026-03-21 | 7/350 |
|