Part One: Linear Algebra ($50$ points)
- ($5\%$) Which of the following statements are true?
- A linear system may have exactly $2$ solutions.
- A linear system of $4$ equations in $3$ unknown will never have a unique solution.
- If $AA^T=0$, then $A=0$.
- $(A+B)^T=A^{-1}+B^{-1}$.
- If $A$ is a squre matrix, then $A+A^T$ is symmetric.
- 若一個線性方程組有兩解,則可以推出有無窮多解,故不可能恰有兩解。更明確的說,若 $x_1\neq x_2$ 為 $Ax=b$ 的解,那麼 $x_1-x_2$ 為 $Ax=0$ 的解,那麼我們可以注意到 $x=x_1+t\left(x_1-x_2\right)$ 皆為 $Ax=b$ 的解,其中 $t\in\mathbb{R}$。因此 $Ax=b$ 不可能恰有兩解,本命題錯誤。
- 本命題未必正確。不妨考慮方程組
$\left\{\begin{aligned}&x=0\\&y=0\\&z=0\\&z=0\end{aligned}\right.$
決定出唯一解 $\left(x,y,z\right)=\left(0,0,0\right)$,故確實有可能有唯一解。 - 本命題未必正確,當 $A$ 為實係數矩陣才正確。反例如下:
$A=\begin{bmatrix}1&i\\1&i\end{bmatrix}\neq0$
但 $AA^T=0$。 - 本命題是錯的。考慮反例:$A=\begin{bmatrix}1&0\\0&1\end{bmatrix}$,$B=\begin{bmatrix}1&0\\1&1\end{bmatrix}$,那麼 $A^{-1}=\begin{bmatrix}1&0\\0&1\end{bmatrix}$,$B^{-1}=\begin{bmatrix}1&0\\-1&1\end{bmatrix}$。因此
$\left(A+B\right)^T=\begin{bmatrix}2&0\\1&2\end{bmatrix}^T=\begin{bmatrix}2&1\\0&2\end{bmatrix}\neq\begin{bmatrix}2&0\\-1&2\end{bmatrix}=A^{-1}+B^{-1}$
- 直接確認可知
$\left(A+A^T\right)^T=A^T+A=A+A^T$
故 $A+A^T$ 為對稱矩陣。 - ($5\%$) Given the following network of one-way street, find the possible flows along each street. (i.e., $f_1,\cdots,f_5$)
- 對上方的節點有:$f_1+f_2=50$;
- 對左方的節點有:$f_1+f_4=f_3+40$;
- 對右方的節點有:$f_2+f_5+f_3=60$;
- 對下方的節點有:$f_4+f_5=50$。
- ($9\%$)
- Let $A,B,C$ denote $n\times n$ matrices and assume that $\det\left(A\right)=-1$, $\det\left(B\right)=2$, and $\det\left(C\right)=3$. Evaluate $\det\left(A^2BC^TB^{-1}\right)$.
- If $A$ is $3\times3$ and $\det\left(2A^{-1}\right)=-4=\det\left(A^3(B^{-1})^T\right)$ find $\det\left(A\right)$ and $\det\left(B\right)$.
- If $\det\begin{bmatrix}a&b\\c&d\end{bmatrix}=-2$, calculate $\det\begin{bmatrix}2b&0&4d\\1&2&-2\\a+1&2&2\left(c-1\right)\end{bmatrix}=$?
- 運用行列式的公式計算即可
$\displaystyle\det\left(A^2BC^TB^{-1}\right)=\left[\det\left(A\right)\right]^2\det\left(B\right)\det\left(C\right)\det\left(B^{-1}\right)=\left(-1\right)^2\cdot2\cdot3\cdot\frac12=3$
- 首先注意到
$-4=\det\left(2A^{-1}\right)=2^3\left[\det\left(A\right)\right]^{-1}$
故 $\displaystyle\det\left(A\right)=-2$。另一方面,也有$-4=\det\left(A^3\left(B^{-1}\right)^T\right)=\left(\det\left(A\right)\right)^3\left[\det\left(B\right)\right]^{-1}=\left(-2\right)^3\left[\det\left(B\right)\right]^{-1}$
因此解得 $\det\left(B\right)=2$。 - 由條件知 $ad-bc=-2$,而直接計算三階行列式有
$\det\begin{bmatrix}2b&0&4d\\1&2&-2\\a+1&2&2\left(c-1\right)\end{bmatrix}=8b\left(c-1\right)+0+8d-8d\left(a+1\right)+8b-0=-8\left(ad-bc\right)=-8\cdot\left(-2\right)=16$
- ($6\%$) A small town has three primary industries: a copper mine, a railroad, and an electric utility. To produce $\$1$ of copper, the copper mine uses $\$0.20$ of copper, $\$0.10$ of transportation, and $\$0.20$ of electric power. To provide $\$1$ of transportation, the railroad uses $\$0.10$ of copper, $\$0.10$ of transportation, and $\$0.40$ of electric power. To provide $\$1$ of electric power, the electric utility uses $\$0.20$ of copper, $\$0.20$ of transportation, and $\$0.30$ of electric power. Suppose that there is an outside demand of $\$12$, $\$8$ and $\$15$ for copper, transportation, and electric power respectively. How much should each industry produce?
- ($12\%$) Answer each of the following as true (T) or false (F).
- Let $L:\mathbb{R}^n\to\mathbb{R}^n$ be a linear transformation defined by $L(X)=AX$ for $X$ in $\mathbb{R}^n$. Then $\dim$ (range $L$) $=n$ if and only if rank $A=n$.
- The set of all vectros of the form $\left(a,b,-a\right)$ is a subspace of $\mathbb{R}^3$.
- Span$\left\{\left(1,1,0\right),\left(0,1,1\right),\left(1,0,1\right)\right\}=\mathbb{R}^3$.
- Every set of $5$ orthonormal vectors is a basis for $\mathbb{R}^5$.
- Let $\mathbb{P}_2\to\mathbb{P}_2$ be defined as $L\left(at^2+bt+c\right)=2at-b$, then $L$ is a linear transformation.
- A Markov chain will have a unique steady-state vector only if its transition matrix is regular.
- 矩陣的秩(rank)便表示行空間與列空間的維度,而矩陣的行空間便是該線性變換的值域,故兩者維數必定相同,本選項正確。
- 本選項正確。設 $V=\left\{\left(a,b,-a\right)\in\mathbb{R}^3:~a,b\in\mathbb{R}\right\}$,容易觀察到
- 原點 $\left(0,0,0\right)\in V$;
- 給定 $v_1=\left(a_1,b_1,-a_1\right)$ 與 $v_2=\left(a_2,b_2,-a_2\right)$ 以及實數 $c\in\mathbb{R}$,有
$v_1+cv_2=\left(\left(a_1+ca_2\right),b_1+cb_2,-\left(a_1+ca_2\right)\right)\in V$
- 本選項正確。考慮一個五元正交集 $\mathcal{B}=\left\{v_1,v_2,v_3,v_4,v_5\right\}$,並給定一項量 $v\in\mathbb{R}^5$,那麼容易看出
$\displaystyle v=\sum_{k=1}^5\langle v,v_k\rangle v_k$
此處 $\langle~,~\rangle$ 表 $\mathbb{R}^5$ 中通常的內積。上式表明 $\mathcal{B}$ 能夠生成 $\mathbb{R}^5$。又因 $\mathcal{B}$ 為五元集,故知此集合能作為基底。 - 本選項正確。任意給定 $\mathbb{R}^3$ 中的向量 $\left(a,b,c\right)$,可以注意到
$\displaystyle\left(a,b,c\right)=\frac{a+b-c}2\left(1,1,0\right)+\frac{-a+b+c}2\left(0,1,1\right)+\frac{a-b+c}2\left(1,0,1\right)$
故 Span$\left\{\left(1,1,0\right),\left(0,1,1\right),\left(1,0,1\right)\right\}=\mathbb{R}^3$。 - 本選項正確。給定實數 $d$ 可直接檢驗發現到
$\begin{aligned}L\left(\left(a_1t^2+b_1t+c_1\right)+d\left(a_2t^2+b_2t+c_2\right)\right)&=L\left(\left(a_1+da_2\right)t^2+\left(b_1+db_2\right)t+\left(c_1+dc_2\right)\right)\\&=2\left(a_1+da_2\right)t-\left(b_1+db_2\right)\\&=\left(2a_1t-b_1\right)+d\left(2a_2t-b_2\right)\\&=L\left(a_1t^2+b_1t+c_1\right)+dL\left(a_2t^2+b_2t+c_2\right)\end{aligned}$
- 本命題不成立。例如轉移矩陣 $A=\begin{bmatrix}1&1\\0&0\end{bmatrix}$ 有唯一的穩定態 $\begin{bmatrix}1\\0\end{bmatrix}$,但 $A$ 並非正規矩陣。
- ($4\%$) Which of the following are equivalent to that $A$ is nonsingular?
- $AX=0$ has infinite solutions.
- $A$ is row equivalent to $I_n$.
- The linear system has a unique solution for every $n\times1$ matrix $B$.
- $\det A\neq0$.
- If $A^2=A$, then $\det A=1$.
- 本選項錯誤。若 $A$ 為非奇異矩陣,那麼 $AX=0$ 僅有零解,而非無窮多解。
- 本選項正確,若矩陣 $A$ 非奇異,那麼便存在一系列列運算化簡為單位方陣;反之亦然。
- 本選項正確。由於對於每個 $n\times1$ 的矩陣 $B$ 皆有解且唯一解,這表明 $A$ 非奇異;反之亦然。
- 本選項正確。矩陣非奇異等同矩陣存在反方陣,因此等價於 $\det A\neq0$。
- 本命題錯誤。考慮反例 $A=\begin{bmatrix}0&1\\0&0\end{bmatrix}$,此使敘述「若 $A^2=A$,則 $\det A=1$」的前提假且結論假,故敘述為真,但明顯 $A$ 為奇異矩陣。
- ($9\%$) Let $A=\begin{bmatrix}1&0&0\\1&2&-3\\1&-1&0\end{bmatrix}$.
- Find the eigenvalues for $A$. Is $A$ is diagonalizable?
- What is the value of trace$(4A)$?
- Find one corresponding eigenvector for any eigenvalue.
- 矩陣 $A$ 所對應的特徵多項式為
$\det\left(xI_3-A\right)=\det\begin{bmatrix}x-1&0&0\\-1&x-2&3\\-1&1&x\end{bmatrix}=\left(x-1\right)\left[x\left(x-2\right)-3\right]=\left(x+1\right)\left(x-1\right)\left(x-3\right)$
由於 $3$ 階方陣有三個相異的特徵值 $\pm1$ 與 $3$,故可對角化。 - 按跡的定義可知
trace$\left(4A\right)=$trace$\begin{bmatrix}4&0&0\\4&8&-12\\4&-4&0\end{bmatrix}=4+8+0=12$
亦可留意到 $4A$ 的特徵值分別為 $\pm4$ 與 $12$,而跡則為特徵值之和故為 $12$。 - 特徵向量分別計算如下
- 若 $x=-1$,則矩陣 $-I_3-A=\begin{bmatrix}-2&0&0\\-1&-3&3\\-1&1&-1\end{bmatrix}$,可取 $v_1=\begin{bmatrix}0\\1\\1\end{bmatrix}\in N\left(-I_3-A\right)$,可以檢驗有 $Av_1=-v_1$。
- 若 $x=1$,則矩陣 $I_3-A=\begin{bmatrix}0&0&0\\-1&-1&3\\-1&1&1\end{bmatrix}$,可取 $v_2=\begin{bmatrix}2\\1\\1\end{bmatrix}\in N\left(I_3-A\right)$,可以檢驗有 $Av_2=v_2$。
- 若 $x=3$,則矩陣 $3I_3-A=\begin{bmatrix}2&0&0\\-1&1&3\\-1&1&3\end{bmatrix}$,可取 $v_3=\begin{bmatrix}0\\3\\-1\end{bmatrix}\in N\left(3I_3-A\right)$,可以檢驗有 $Av_3=3v_3$。
訣竅
經由一般計算的經驗可容易舉出反例,而正確的命題則亦可給出計算式的證明。解法
訣竅
依據克希荷夫原理(Kirchhoffs Priciple)[每個節點所流入的量與流出的量相同]來列式並求解即可。解法
依據每個節點來列式如下$\begin{bmatrix}1&1&0&0&0\\1&0&-1&1&0\\0&1&1&0&1\\0&0&0&1&1\end{bmatrix}\begin{bmatrix}f_1\\f_2\\f_3\\f_4\\f_5\end{bmatrix}=\begin{bmatrix}50\\40\\60\\50\end{bmatrix}$
使用列運算計算至最簡列梯形式可得$\begin{bmatrix}1&0&-1&0&-1\\0&1&1&0&1\\0&0&0&1&1\\0&0&0&0&0\end{bmatrix}\begin{bmatrix}f_1\\f_2\\f_3\\f_4\\f_5\end{bmatrix}=\begin{bmatrix}-10\\60\\50\\0\end{bmatrix}$
設 $f_3=c_1$,$f_5=c_2$,那麼可解得$\begin{bmatrix}f_1\\f_2\\f_3\\f_4\\f_5\end{bmatrix}=\begin{bmatrix}-10\\60\\0\\50\\0\end{bmatrix}+c_1\begin{bmatrix}1\\-1\\1\\0\\0\end{bmatrix}+c_2\begin{bmatrix}1\\-1\\0\\-1\\1\end{bmatrix}$
其中變數 $c_1$ 與 $c_2$ 滿足 $\left\{\begin{aligned}&10\leq c_1+c_2\leq60\\&0\leq c_2\leq50\\&0\leq c_1\end{aligned}\right.$。訣竅
運用行列式的性質求解。解法
訣竅
依據題意列式即可求解。解法
設定內部消耗矩陣 $C$ 如下$\displaystyle C=\begin{bmatrix}0.2&0.1&0.2\\0.1&0.1&0.2\\0.2&0.4&0.3\end{bmatrix}$
以及需求 $D=\begin{bmatrix}12\\8\\15\end{bmatrix}$。若工廠應分別生產 $x$ 單位的銅、$y$ 單位的運輸與 $z$ 單位的電力,那麼應滿足下列關係式$\begin{bmatrix}0.8&-0.1&-0.2\\-0.1&0.9&-0.2\\-0.2&-0.4&0.7\end{bmatrix}\begin{bmatrix}x\\y\\z\end{bmatrix}=\left(I-C\right)X=X-CX=D=\begin{bmatrix}12\\8\\15\end{bmatrix}$
故所求為$\displaystyle X=\begin{bmatrix}x\\y\\z\end{bmatrix}=\begin{bmatrix}0.8&-0.1&-0.2\\-0.1&0.9&-0.2\\-0.2&-0.4&0.7\end{bmatrix}^{-1}\begin{bmatrix}12\\8\\15\end{bmatrix}=\frac1{77}\begin{bmatrix}110&30&40\\22&104&36\\44&68&142\end{bmatrix}\begin{bmatrix}12\\8\\15\end{bmatrix}=\frac1{77}\begin{bmatrix}2160\\1636\\3202\end{bmatrix}$
訣竅
依線性代數的基本知識即可答題。解法
訣竅
對於非奇異矩陣的充分且必要條件有基本的認識。解法
訣竅
按特徵多項式的定義求其特徵值與特徵向量,從而判定可否對角化;按照跡(trace)的定義計算即可。解法
【註】取矩陣
$P=\begin{bmatrix}v_1&v_2&v_3\end{bmatrix}=\begin{bmatrix}0&2&0\\1&1&3\\1&1&-1\end{bmatrix},\quad \Lambda=\begin{bmatrix}-1&0&0\\0&1&0\\0&0&3\end{bmatrix}$
那麼容易檢驗$A=P\Lambda P^{-1}$
此即矩陣 $A$ 相似於對角矩陣 $\Lambda$,故直接說明了 $A$ 可對角化。Part Two: Calculus ($50$ points)
- ($5$ points) Suppose that $f$ is a real function such that $f'\left(x_0\right)$ exists. Is
$\displaystyle\lim_{h\to0}\frac{f\left(x_0+h\right)-f\left(x_0-h\right)}h=f'\left(x_0\right)$?
Why or why not? - ($5$ points)
$\displaystyle\frac{d}{dx}\int_0^{x^2}e^{-t^2}dt=$?
- ($6$ points) If $\displaystyle f\left(x\right)=\sum_{n=0}^{\infty}\left(-1\right)^nx^{2n}$ for all $x\in\left(0,1\right)$, then $f'\left(x\right)=$?
- ($6$ points)
$\displaystyle\lim_{n\to\infty}\frac3n\sum_{i=1}^n\left\{\left(\frac{3i}n\right)^2-\left(\frac{3i}n\right)\right\}=$?
- ($6$ points) Compute $\displaystyle\int_{\alpha}\left[\left(x+y\right)dx+\left(x-y\right)dy\right]$, where $\alpha$ is any smooth curve joining the points $\left(0,1\right)$ and $\left(2,3\right)$.
- ($6$ points) If
$\displaystyle f\left(x,y\right)=\frac{2y-1}{x+1}$,
and if the region $T$ is bounded by $x=0$, $y=0$, and $2x-y-4=0$, find the double integral of $f$ over $T$. - ($6$ points) Find the volume of the solid $S$ of intersection of two right circular cylinders of radius $r$, assuming that their axes meet at right angle.
- ($10$ points) Suppose we are hiring a weather forecaster to predict the probability that next summer will be rainy or sunny. The following suggests a method that can be used to ensure that the forecaster is accurate. Suppose that the actual probability of rain next summer is $q$. For simplicity, we assume that the summer can only be rainy or sunny. If the forecaster announces a probability $p$ that the summer will be rainy, he or she receives a payment of $1-\left(1-p\right)^2$ if the summer is rainy and a payment of $1-p^2$ if the summer is sunny. Show that the forecaster will maximize expected profits by announcing that the probability of a rainy summer is $q$.
訣竅
運用導數的定義求解即可。解法
改寫極限式注意到為$\displaystyle\lim_{h\to0}\frac{f\left(x_0+h\right)-f\left(x_0-h\right)}h=\lim_{h\to0}\frac{f\left(x_0+h\right)-f\left(x_0\right)}h+\lim_{h\to0}\frac{f\left(x_0-h\right)-f\left(x_0\right)}{-h}=2f'\left(x_0\right)$
一般而言不會等於 $f'\left(x_0\right)$ 除非 $f'\left(x_0\right)=0$,故該敘述未必正確。訣竅
運用微積分基本定理與連鎖律直接計算即可。解法
運用微積分基本定理與連鎖律直接計算可知$\displaystyle\frac{d}{dx}\int_0^{x^2}e^{-t^2}dt=e^{-\left(x^2\right)^2}\cdot2x=2xe^{-x^4}$
訣竅
將函數 $f$ 視為無窮等比級數和,隨後求導。解法
由於 $x\in\left(0,1\right)$,因此$-x^2\in\left(-1,0\right)$,因此$\displaystyle f\left(x\right)=\sum_{n=0}^{\infty}\left(-1\right)^nx^{2n}=\sum_{n=0}^{\infty}\left(-x^2\right)^n=\frac1{1+x^2}~~\mbox{for}~x\in\left(0,1\right)$
因此求導可得$\displaystyle f'\left(x\right)=\sum_{n=0}^{\infty}2n\left(-1\right)^nx^{2n-1}=-\frac{2x}{\left(1+x^2\right)^2}~~\mbox{for}~x\in\left(0,1\right)$
訣竅
可將此視為黎曼和,取極限後化為定積分計算之;亦可求其和後直接計算極限。解法一
設 $f\left(x\right)=x^2-x$,那麼該極限式可視為 $f$ 在 $\left[0,3\right]$ 上的上黎曼和,故取極限可計算如下$\displaystyle\int_0^3f\left(x\right)dx=\int_0^3\left(x^2-x\right)dx=\left.\left(\frac{x^3}3-\frac{x^2}2\right)\right|_0^3=\frac92$
解法二
直接求和可知$\displaystyle\begin{aligned}\sum_{i=1}^n\left\{\left(\frac{3i}n\right)^2-\left(\frac{3i}n\right)\right\}&=\frac9{n^2}\sum_{i=1}^ni^2-\frac3n\sum_{i=1}^ni=\frac9{n^2}\cdot\frac{n\left(n+1\right)\left(2n+1\right)}6-\frac3n\cdot\frac{n\left(n+1\right)}2\\&=\frac{3\left(n+1\right)\left(2n+1\right)}{2n}-\frac{3n+3}2=\frac{3\left(n+1\right)^2}{2n}\end{aligned}$
因此所求為$\displaystyle\lim_{n\to\infty}\frac3n\sum_{i=1}^n\left\{\left(\frac{3i}n\right)^2-\left(\frac{3i}n\right)\right\}=\lim_{n\to\infty}\frac{9\left(n+1\right)^2}{2n^2}=\frac92$
訣竅
運用線積分基本定理即可。解法
運用偏積分觀察可取$\displaystyle f\left(x,y\right)=\frac{x^2+2xy-y^2}2+C$
那麼所求為$\displaystyle\int_{\alpha}\left[\left(x+y\right)dx+\left(x-y\right)dy\right]=\int_{\alpha}\nabla f\cdot d\vec{r}=f\left(2,3\right)-f\left(0,1\right)=\left(\frac72+C\right)-\left(-\frac12+C\right)=4$
訣竅
明確地使用不等式表達出區域 $T$,隨後運用 Fubini 定理表達該重積分計算,而對於雙變數的順序選擇導致兩種不同的計算過程。解法一
積分區域 $T$ 可表達為 $\left\{\begin{aligned}&0\leq x\leq2\\&2x-4\leq y\leq0\end{aligned}\right.$,如此所求的重積分可表達並計算如下$\displaystyle\begin{aligned}\iint_Tf\left(x,y\right)dA&=\int_0^2\int_{2x-4}^0\frac{2y-1}{x+1}dydx=\int_0^2\left.\frac{y^2-y}{x+1}\right|_{y=2x-4}^{y=0}dx\\&=-2\int_0^2\frac{2x^2-9x+10}{x+1}dx=-2\int_0^2\left(2x-11+\frac{21}{x+1}\right)dx\\&=\left.-2\left(x^2-11x+21\ln\left(x+1\right)\right)\right|_0^2=36-42\ln3\end{aligned}$
解法二
積分區域 $T$ 可表達為 $\left\{\begin{aligned}&0\leq x\leq\left(y+4\right)/2\\&-4\leq y\leq0\end{aligned}\right.$,如此所求的重積分可表達並計算如下$\displaystyle\begin{aligned}\iint_Tf\left(x,y\right)dA&=\int_{-4}^0\int_0^{\frac{y+4}2}\frac{2y-1}{x+1}dxdy=\int_{-4}^0\left(2y-1\right)\ln\left(x+1\right)\Big|_{x=0}^{x=\frac{y+4}2}dy\\&=\int_{-4}^0\left[\left(2y-1\right)\ln\left(y+6\right)-\left(2y-1\right)\ln2\right]dy\\&=\int_{-4}^0\ln\left(y+6\right)d\left(y^2-y\right)-\ln2\int_{-4}^0\left(2y-1\right)dy\\&=\left(y^2-y\right)\ln\left(y+6\right)\Big|_{-4}^0-\ln2\left(y^2-y\right)\Big|_{-4}^0-\int_{-4}^0\frac{y^2-y}{y+6}dy\\&=-\int_{-4}^0\left[y-7+\frac{42}{y+6}\right]dy=\left.-\left(\frac{y^2}2-7y+42\ln\left(y+6\right)\right)\right|_{-4}^0\\&=36-42\ln3\end{aligned}$
訣竅
按照題意的情境做出適當的假設,隨後選定一平面作為底面後,考慮由底面積成以高的概念出發以重積分的形式列出體積算式並計算。解法
考慮兩半徑為 $r$ 的垂直圓柱$C_1=\left\{\left(x,y,z\right)\in\mathbb{R}^3:x^2+y^2\leq r^2\right\}$
以及$C_2=\left\{\left(x,y,z\right)\in\mathbb{R}^3:x^2+z^2\leq r^2\right\}$
並且考慮其交集 $V=C_1\cap C_2$,而 $C_1$ 投影至 $xy$ 平面的集合為$D=\left\{\left(x,y\right)\in\mathbb{R}^2:x^2+y^2\leq r^2\right\}$
那麼以 $D$ 為底面而以 $z=\pm\sqrt{r^2-x^2}$ 分別為上下曲面可列出體積算式如下$\displaystyle\left|V\right|=\iint_D2\sqrt{r^2-x^2}dA$
運用極座標變換,令 $\left\{\begin{aligned}&x=\rho\cos\theta\\&y=\rho\sin\theta\end{aligned}\right.$,其中變數範圍為 $\left\{\begin{aligned}&0\leq\rho\leq r\\&0\leq\theta\leq2\pi\end{aligned}\right.$,故所求的體積可改寫並計算如下$\displaystyle\begin{aligned}\left|V\right|&=2\int_0^{2\pi}\int_0^r\sqrt{r^2-\rho^2\cos^2\theta}\rho d\rho d\theta=8\int_0^{\frac\pi2}\int_0^r\sqrt{r^2-\rho^2\cos^2\theta}\rho d\rho d\theta\\&=-\frac83\int_0^{\frac\pi2}\left.\frac{\left(r^2-\rho^2\cos^2\theta\right)^{3/2}}{\cos^2\theta}\right|_{\rho=0}^{\rho=r}d\theta=\frac{8r^3}3\int_0^{\frac\pi2}\sec^2\theta\left(1-\sin^3\theta\right)d\theta\\&=\left.\frac{8r^3}3\left(\tan\theta-\sec\theta-\cos\theta\right)\right|_0^{\frac\pi2}=-\left.\frac{8r^3}3\left(\frac1{\tan\theta+\sec\theta}+\cos\theta\right)\right|_0^{\frac\pi2}\\&=\frac{16r^3}3\end{aligned}$
訣竅
依據題意列出天氣預測者的報償函數,其中變量為他或她所宣告的降雨率 $p$。我們利用微分求導使報償函數之值達到最大,進而發現 $p=q$ 時可達到最大值。解法
考慮這位天氣預測者的報償函數如下$P\left(p\right)=q\cdot\left(1-\left(1-p\right)^2\right)+\left(1-q\right)\cdot\left(1-p^2\right)=1+2pq-p^2-q$
其中 $0\leq p\leq1$。現在將 $P$ 對 $p$ 求導可知$P'\left(p\right)=2q-2p$
故 $P'\left(q\right)=0$。再者 $P''\left(p\right)=-2<0$,因此僅當 $p=q$ 時天氣預測者的報償可達到最大值,故此人欲使自身獲得的報酬提高,那他必然會盡可能使 $p$ 值盡可能接近 $q$ 值。[即便 $q=0$ 或 $q=1$ 時亦同。]
沒有留言:
張貼留言