| 24小時(shí)熱門(mén)版塊排行榜 |
| 9 | 1/1 | 返回列表 |
| 查看: 3541 | 回復(fù): 8 | |||
cailianguo金蟲(chóng) (正式寫(xiě)手)
|
[交流]
擬合和回歸有什么區(qū)別
|
||
| 如題,擬合和回歸有什么區(qū)別與聯(lián)系?請(qǐng)賜教! |
» 搶金幣啦!回帖就可以得到:
+1/633
+1/461
+5/195
+1/194
+5/175
+1/90
+2/84
+1/84
+1/82
+1/79
+1/78
+1/75
+1/15
+1/15
+1/14
+1/6
+1/4
+1/4
+2/4
+1/2
至尊木蟲(chóng) (正式寫(xiě)手)
銀蟲(chóng) (小有名氣)
金蟲(chóng) (初入文壇)
|
擬合是一種數(shù)據(jù)處理的方式。簡(jiǎn)單的說(shuō)就是你有一組數(shù)據(jù),覺(jué)得這組數(shù)據(jù)和一個(gè)已知的函數(shù)(這個(gè)函數(shù)的參數(shù)未定)很相似,為了得到最能表示這組數(shù)據(jù)特征的這個(gè)函數(shù),通過(guò)擬合這種方式(具體的數(shù)學(xué)方法很多)求得參數(shù)。 而回歸是一種特定的數(shù)學(xué)方法,它可以實(shí)現(xiàn)數(shù)據(jù)擬合,得到函數(shù)的參數(shù)。 也有些擬合得到的參數(shù)并非是函數(shù)的參數(shù),如神經(jīng)網(wǎng)絡(luò),得到的是這個(gè)神經(jīng)網(wǎng)絡(luò)的參數(shù)。 |
木蟲(chóng) (著名寫(xiě)手)
至尊木蟲(chóng) (文壇精英)
金蟲(chóng) (職業(yè)作家)
|
這是網(wǎng)上對(duì)擬合、差值、逼近、回歸這些概念的分析 1、回歸一般指線(xiàn)性回歸,是求最小二乘解的過(guò)程; 2、擬合包括插值與逼近; 3、插值曲線(xiàn)要經(jīng)過(guò)型值點(diǎn); 4、逼近只要求曲線(xiàn)接近型值點(diǎn),符合型值點(diǎn)趨勢(shì)。 A、在求回歸前,已經(jīng)假設(shè)所有型值點(diǎn)同時(shí)滿(mǎn)足某一曲線(xiàn)方程,計(jì)算只要求出該方程的系數(shù); B、插值和逼近的結(jié)果曲線(xiàn)方程是由型值點(diǎn)而決定,不是一個(gè)求系數(shù)的過(guò)程。一般來(lái)說(shuō),型值點(diǎn)與型值點(diǎn)之間的曲線(xiàn)方程與鄰近的幾個(gè)型值點(diǎn)關(guān)系最大;離之越遠(yuǎn),關(guān)系越小。 簡(jiǎn)單點(diǎn)說(shuō)吧,插值和擬合差不多,只不過(guò)插值的時(shí)候已知點(diǎn)全部在曲線(xiàn)上(可以證明一定存在這樣的曲線(xiàn)),擬合就不一定要已知點(diǎn)全部在曲線(xiàn)上了;貧w分析其實(shí)就是插值、擬合的進(jìn)一步操作,包括F檢驗(yàn)、殘差分析等東西,也就是在插值擬合的基礎(chǔ)上再做一些分析。這三者沒(méi)有什么本質(zhì)的區(qū)別,樓主不用太擔(dān)心。 通過(guò)上面的分析,我感覺(jué) 擬合的概念更廣泛,回歸是擬合的一種方法。 |
金蟲(chóng) (小有名氣)
|
Curve fitting is the process of constructing a curve, or mathematical function, that has the best fit to a series of data points, possibly subject to constraints. Curve fitting can involve either interpolation, where an exact fit to the data is required, or smoothing, in which a "smooth" function is constructed that approximately fits the data. A related topic is regression analysis, which focuses more on questions of statistical inference such as how much uncertainty is present in a curve that is fit to data observed with random errors. Fitted curves can be used as an aid for data visualization, to infer values of a function where no data are available, and to summarize the relationships among two or more variables. Extrapolation refers to the use of a fitted curve beyond the range of the observed data, and is subject to a greater degree of uncertainty since it may reflect the method used to construct the curve as much as it reflects the observed data. cite:http://blog.sina.com.cn/s/blog_53b8f53e0102e1pw.html ,我覺(jué)得這個(gè)說(shuō)的很明確 |
| 9 | 1/1 | 返回列表 |
| 最具人氣熱帖推薦 [查看全部] | 作者 | 回/看 | 最后發(fā)表 | |
|---|---|---|---|---|
|
[考研] 333求調(diào)劑 +6 | 87639 2026-03-21 | 10/500 |
|
|---|---|---|---|---|
|
[考研]
|
2117205181 2026-03-21 | 8/400 |
|
|
[考研] 一志愿西安交通大學(xué)材料工程專(zhuān)業(yè) 282分求調(diào)劑 +11 | 楓橋ZL 2026-03-18 | 13/650 |
|
|
[考研] 材料與化工085600,總分304,本科有兩篇sci參與,求調(diào)劑 +4 | 幸運(yùn)的醬醬 2026-03-22 | 5/250 |
|
|
[考研] 289材料與化工(085600)B區(qū)求調(diào)劑 +3 | 這么名字咋樣 2026-03-22 | 4/200 |
|
|
[考研] 生物學(xué)071000 329分求調(diào)劑 +5 | 我愛(ài)生物生物愛(ài)?/a> 2026-03-17 | 5/250 |
|
|
[考研] 384求調(diào)劑 +3 | 子系博 2026-03-22 | 4/200 |
|
|
[考研] 286分人工智能專(zhuān)業(yè)請(qǐng)求調(diào)劑愿意跨考! +4 | lemonzzn 2026-03-17 | 8/400 |
|
|
[考研] 求調(diào)劑 +3 | 13341 2026-03-20 | 3/150 |
|
|
[考研] 求調(diào)劑 +6 | Mqqqqqq 2026-03-19 | 6/300 |
|
|
[考研] 材料工程(專(zhuān))一志愿985 初試335求調(diào)劑 +3 | hiloiy 2026-03-17 | 4/200 |
|
|
[考研] 317求調(diào)劑 +5 | 申子申申 2026-03-19 | 9/450 |
|
|
[考研] 中南大學(xué)化學(xué)學(xué)碩337求調(diào)劑 +3 | niko- 2026-03-19 | 6/300 |
|
|
[考研] 261求B區(qū)調(diào)劑,科研經(jīng)歷豐富 +3 | 牛奶很忙 2026-03-20 | 4/200 |
|
|
[考研] 求調(diào)劑 +3 | eation27 2026-03-20 | 3/150 |
|
|
[考研] 廣西大學(xué)家禽遺傳育種課題組2026年碩士招生(接收計(jì)算機(jī)專(zhuān)業(yè)調(diào)劑) +3 | 123阿標(biāo) 2026-03-17 | 3/150 |
|
|
[考研] 材料學(xué)碩318求調(diào)劑 +5 | February_Feb 2026-03-19 | 5/250 |
|
|
[考研] 288求調(diào)劑,一志愿華南理工大學(xué)071005 +5 | ioodiiij 2026-03-17 | 5/250 |
|
|
[考研] 085600材料與化工求調(diào)劑 +6 | 緒幸與子 2026-03-17 | 6/300 |
|
|
[考研] 考研調(diào)劑 +3 | 淇ya_~ 2026-03-17 | 5/250 |
|