๋นˆ๋„ ๊ฐ€์ค‘ ab-์ง€์ˆ˜ ํšŒ๊ท€ ๊ณ„์‚ฐ๊ธฐ (y = A·B^x)

๋นˆ๋„ ๊ฐ€์ค‘ ab-์ง€์ˆ˜ ํšŒ๊ท€ ๊ณ„์‚ฐ๊ธฐ (y = A·B^x)

์‹ ํฅ AI Skill PRO

๊ณ„์‚ฐ ์ž…๋ ฅ

๊ฐ ์ค„์€ ํ•˜๋‚˜์˜ ํ–‰์ž…๋‹ˆ๋‹ค: x ๊ฐ’, y ๊ฐ’(๋ฐ˜๋“œ์‹œ > 0), ๊ทธ๋ฆฌ๊ณ  ๋นˆ๋„ f(ํšŸ์ˆ˜/๊ฐ€์ค‘์น˜). ๊ฐ€์ค‘์น˜ ์—†๋Š” ์ผ๋ฐ˜ ์ ํ•ฉ์„ ์›ํ•˜๋ฉด ๋ชจ๋“  f๋ฅผ 1๋กœ ์„ค์ •ํ•˜์„ธ์š”.

๊ณต์‹

y=AยทBx,B=eSxy/Sxx,A=elnโกyโ€พโˆ’xโ€พ(Sxy/Sxx)์—ฌ๊ธฐ์„œ{n=โˆ‘f,xโ€พ=โˆ‘fxn,lnโกyโ€พ=โˆ‘flnโกynSxx=โˆ‘fx2โˆ’nxโ€พ2Sxy=โˆ‘fxlnโกyโˆ’nxโ€พlnโกyโ€พx,y,f=๋ฐ์ดํ„ฐ ์  (x, y, f)\begin{gathered} y = A \cdot B^{x}, \qquad B = e^{\,S_{xy}/S_{xx}}, \quad A = e^{\,\overline{\ln y} - \bar{x}\,(S_{xy}/S_{xx})} \\[1.5em] \text{์—ฌ๊ธฐ์„œ}\quad \left\{ \begin{aligned} n &= \textstyle\sum f, \quad \bar{x} = \frac{\sum f x}{n}, \quad \overline{\ln y} = \frac{\sum f \ln y}{n} \\ S_{xx} &= \textstyle\sum f x^{2} - n\,\bar{x}^{2} \\ S_{xy} &= \textstyle\sum f\,x \ln y - n\,\bar{x}\,\overline{\ln y} \\ x,\,y,\,f &= \text{๋ฐ์ดํ„ฐ ์  (x, y, f)} \end{aligned} \right. \end{gathered}
์‚ฌ์šฉ ํ†ต๊ณ„

๊ด‘๊ณ 

๊ฒฐ๊ณผ

์ ํ•ฉ๋œ ab-์ง€์ˆ˜ ์ถ”์„ธ์„ 
y = 1.000000000 ยท 2.000000000^x
y = A ยท Bx
๊ณ„์ˆ˜ A 1
๋ฐ‘ B 2
์ƒ๊ด€๊ณ„์ˆ˜ r 1
์œ ํšจ ํ‘œ๋ณธ ํฌ๊ธฐ n (ฮฃ f) 4

์ด ํŽ˜์ด์ง€์˜ ์˜ค๋ฅ˜ ์‹ ๊ณ 

์ด ๊ณ„์‚ฐ๊ธฐ์˜ ๊ธฐ๋Šฅ

์ด ๋„๊ตฌ๋Š” ๋นˆ๋„ ๋ถ„ํฌํ‘œ ํ˜•ํƒœ์˜ ๋ฐ์ดํ„ฐ์— \(y = A \cdot B^{x}\) ๊ผด์˜ ab-์ง€์ˆ˜ ์ถ”์„ธ์„ ์„ ์ ํ•ฉํ•ฉ๋‹ˆ๋‹ค. ๊ฐ ํ–‰์€ x ๊ฐ’, y ๊ฐ’, ๊ทธ๋ฆฌ๊ณ  ๋นˆ๋„ \(f\)(ํ•ด๋‹น (x, y) ์Œ์ด ๋‚˜ํƒ€๋‚˜๋Š” ํšŸ์ˆ˜ ๋˜๋Š” ๊ฐ€์ค‘์น˜)๋กœ ๊ตฌ์„ฑ๋ฉ๋‹ˆ๋‹ค. ๋‹จ์œ„๋‚˜ ํŠน์ • ๊ตญ๊ฐ€ ๊ทœ์ •์— ์–ฝ๋งค์ด์ง€ ์•Š๋Š” ์ˆœ์ˆ˜ ํ†ต๊ณ„ ๊ธฐ๋ฒ•์œผ๋กœ, ์–ด๋””์„œ๋‚˜ ๋™์ผํ•˜๊ฒŒ ์ ์šฉ๋ฉ๋‹ˆ๋‹ค.

์‚ฌ์šฉ ๋ฐฉ๋ฒ•

ํ•œ ์ค„์— ํ•˜๋‚˜์”ฉ ๋ฐ์ดํ„ฐ ์ ์„ x, y, f ํ˜•์‹์œผ๋กœ ์ž…๋ ฅํ•˜์„ธ์š”. ์ด ๋ฐฉ๋ฒ•์€ y์— ์ž์—ฐ๋กœ๊ทธ๋ฅผ ์ทจํ•˜๋ฏ€๋กœ y ๊ฐ’์€ ๋ฐ˜๋“œ์‹œ ์–‘์ˆ˜์—ฌ์•ผ ํ•ฉ๋‹ˆ๋‹ค. ๊ฐ€์ค‘์น˜๋ฅผ ์ ์šฉํ•˜์ง€ ์•Š๋Š” ์ผ๋ฐ˜ ํšŒ๊ท€๋ฅผ ์›ํ•œ๋‹ค๋ฉด ๋ชจ๋“  ๋นˆ๋„๋ฅผ 1๋กœ ์„ค์ •ํ•˜๋ฉด ๋ฉ๋‹ˆ๋‹ค(์„ธ ๋ฒˆ์งธ ์ˆซ์ž๋ฅผ ์ƒ๋žตํ•˜๋ฉด f = 1๋กœ ๊ฐ„์ฃผํ•ฉ๋‹ˆ๋‹ค). ํ‘œ์‹œํ•  ๊ณ„์ˆ˜์˜ ์œ ํšจ์ˆซ์ž ์ž๋ฆฟ์ˆ˜๋ฅผ ์„ ํƒํ•œ ๋’ค A, B, ์ƒ๊ด€๊ณ„์ˆ˜ r ๊ฐ’์„ ํ™•์ธํ•˜์„ธ์š”.

y์˜ ๋กœ๊ทธ๋ฅผ ์ทจํ•˜๋ฉด ์ง€์ˆ˜ ๊ณก์„ ์ด ์ง์„ ์ด ๋ฉ๋‹ˆ๋‹ค
ln y๋ฅผ ์ทจํ•˜๋ฉด y = AยทB^x๊ฐ€ ์„ ํ˜•ํ™”๋˜์–ด ๊ฐ€์ค‘ ์ตœ์†Œ์ œ๊ณฑ๋ฒ•์„ ์ ์šฉํ•  ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

๊ณต์‹ ์„ค๋ช…

ํ•ต์‹ฌ์€ ๋ชจ๋ธ์„ ์„ ํ˜•ํ™”ํ•˜๋Š” ๊ฒƒ์ž…๋‹ˆ๋‹ค. \(y = A \cdot B^{x}\)์˜ ์–‘๋ณ€์— ๋กœ๊ทธ๋ฅผ ์ทจํ•˜๋ฉด \(\ln y = \ln A + x \cdot \ln B\)๊ฐ€ ๋˜์–ด (x, ln y) ๊ณต๊ฐ„์—์„œ ์ง์„ ์ด ๋ฉ๋‹ˆ๋‹ค. ์—ฌ๊ธฐ์— x์— ๋Œ€ํ•œ ln y์˜ ๋นˆ๋„ ๊ฐ€์ค‘ ์ตœ์†Œ์ œ๊ณฑ ํšŒ๊ท€๋ฅผ ์ˆ˜ํ–‰ํ•ฉ๋‹ˆ๋‹ค.

$$y = A \cdot B^{x}, \qquad B = e^{\,S_{xy}/S_{xx}}, \quad A = e^{\,\overline{\ln y} - \bar{x}\,(S_{xy}/S_{xx})}$$$$\text{where}\quad \left\{ \begin{aligned} n &= \textstyle\sum f, \quad \bar{x} = \frac{\sum f x}{n}, \quad \overline{\ln y} = \frac{\sum f \ln y}{n} \\ S_{xx} &= \textstyle\sum f x^{2} - n\,\bar{x}^{2} \\ S_{xy} &= \textstyle\sum f\,x \ln y - n\,\bar{x}\,\overline{\ln y} \\ x,\,y,\,f &= \text{Data points (x, y, f)} \end{aligned} \right.$$

n = \(\sum f\), ๊ฐ€์ค‘ ํ‰๊ท  \(\bar{x}\)์™€ \(\overline{\ln y}\), ๊ฐ€์ค‘ ํ•ฉ \(S_{xx}\), \(S_{yy}\), \(S_{xy}\)๋ฅผ ์‚ฌ์šฉํ•  ๋•Œ ๊ธฐ์šธ๊ธฐ๋Š” \(S_{xy}/S_{xx}\), ์ ˆํŽธ์€ \(\overline{\ln y} - \bar{x} \cdot \text{๊ธฐ์šธ๊ธฐ}\)์ž…๋‹ˆ๋‹ค. ์ด๋ฅผ ์ง€์ˆ˜ํ™”ํ•˜๋ฉด \(B = e^{\text{๊ธฐ์šธ๊ธฐ}}\), \(A = e^{\text{์ ˆํŽธ}}\)์ด ๋ฉ๋‹ˆ๋‹ค. ์ƒ๊ด€๊ณ„์ˆ˜ \(r = S_{xy} / (\sqrt{S_{xx}} \cdot \sqrt{S_{yy}})\)๋Š” ๋กœ๊ทธ-์„ ํ˜• ๋ชจ๋ธ์ด ๋ฐ์ดํ„ฐ์— ์–ผ๋งˆ๋‚˜ ์ž˜ ๋งž๋Š”์ง€๋ฅผ ๋‚˜ํƒ€๋ƒ…๋‹ˆ๋‹ค.

๊ด‘๊ณ 

๊ณ„์‚ฐ ์˜ˆ์‹œ

์ •ํ™•ํžˆ 2๋ฐฐ์”ฉ ์ฆ๊ฐ€ํ•˜๋Š” ๋ฐ์ดํ„ฐ (1, 2), (2, 4), (3, 8), (4, 16)์ด ๋ชจ๋‘ f = 1์ธ ๊ฒฝ์šฐ: n = 4, \(\bar{x} = 2.5\), \(\overline{\ln y} = 1.732868\), \(S_{xx} = 5\), \(S_{xy} = 3.465735\), \(S_{yy} = 2.402224\)์ž…๋‹ˆ๋‹ค. ๋”ฐ๋ผ์„œ $$B = e^{3.465735/5} = e^{0.693147} = 2, \quad A = e^{0} = 1, \quad r = 1$$์ด ๋ฉ๋‹ˆ๋‹ค. ์ฆ‰ ์ ํ•ฉ๋œ ๊ณก์„ ์€ \(y = 1 \cdot 2^{x}\)๋กœ, ์˜ˆ์ƒ๊ณผ ์ •ํ™•ํžˆ ์ผ์น˜ํ•ฉ๋‹ˆ๋‹ค.

๊ฐ€์ค‘ ์ ๋“ค์˜ ์‚ฐ์ ๋„์™€ ์ง€์ˆ˜ ์ตœ์  ์ ํ•ฉ ๊ณก์„ 
๊ฐ€์ค‘ ์ง€์ˆ˜ ์ ํ•ฉ y = AยทB^x: ์ ์ด ํด์ˆ˜๋ก ๊ณก์„ ์—์„œ ๋นˆ๋„ ๊ฐ€์ค‘์น˜๊ฐ€ ํฝ๋‹ˆ๋‹ค.

์ž์ฃผ ๋ฌป๋Š” ์งˆ๋ฌธ

๋นˆ๋„ ์—ด์€ ๋ฌด์Šจ ์—ญํ• ์„ ํ•˜๋‚˜์š”? ๊ฐ ๋ฐ์ดํ„ฐ ์ ์ด ๋‚˜ํƒ€๋‚˜๋Š” ํšŸ์ˆ˜๋งŒํผ ๊ฐ€์ค‘์น˜๋ฅผ ๋ถ€์—ฌํ•ฉ๋‹ˆ๋‹ค. ์ฆ‰ f = 5์ธ ํ–‰์€ ๋™์ผํ•œ ํ–‰ ๋‹ค์„ฏ ๊ฐœ์™€ ๊ฐ™์€ ํšจ๊ณผ๋ฅผ ๋ƒ…๋‹ˆ๋‹ค. ์ผ๋ฐ˜ ํšŒ๊ท€๋ฅผ ์›ํ•œ๋‹ค๋ฉด ๋ชจ๋“  f๋ฅผ 1๋กœ ๋‘์„ธ์š”.

r์€ ์–ด๋–ป๊ฒŒ ํ•ด์„ํ•˜๋‚˜์š”? |r| > 0.7์ด๋ฉด ๊ฐ•ํ•œ ์ƒ๊ด€๊ด€๊ณ„, 0.4โ€“0.7์ด๋ฉด ์ค‘๊ฐ„, 0.2โ€“0.4์ด๋ฉด ์•ฝํ•จ, 0.2 ๋ฏธ๋งŒ์ด๋ฉด ์‚ฌ์‹ค์ƒ ์ƒ๊ด€๊ด€๊ณ„๊ฐ€ ์—†์Œ์„ ์˜๋ฏธํ•ฉ๋‹ˆ๋‹ค. r์€ (x, ln y) ๊ณต๊ฐ„์—์„œ ๊ณ„์‚ฐ๋ฉ๋‹ˆ๋‹ค.

์™œ y๋Š” ์–‘์ˆ˜์—ฌ์•ผ ํ•˜๋‚˜์š”? ์ด ๋ชจ๋ธ์€ y์— ์ž์—ฐ๋กœ๊ทธ๋ฅผ ์ทจํ•˜๋Š”๋ฐ, y โ‰ค 0์ผ ๋•Œ๋Š” ln y๊ฐ€ ์ •์˜๋˜์ง€ ์•Š์Šต๋‹ˆ๋‹ค. ๋”ฐ๋ผ์„œ ์–‘์ˆ˜๊ฐ€ ์•„๋‹Œ ํ–‰์€ ๋ฌด์‹œ๋ฉ๋‹ˆ๋‹ค.

์ตœ์ข… ์—…๋ฐ์ดํŠธ:

์ˆ˜ํ•™ ๋ฐ ํ†ต๊ณ„ ์ธ๊ธฐ ๊ณ„์‚ฐ๊ธฐ

์ˆ˜ํ•™ ๋ฐ ํ†ต๊ณ„ ๊ณ„์‚ฐ๊ธฐ ์ „์ฒด ๋ณด๊ธฐ