What would you like to understand?

Although both graphs curve upward (indicating a positive second derivative), the narrow parabola has a much more rapid increase in slope around $x=4$ compared to the wider parabola. Since the second derivative measures the rate of change of the slope, a faster change in slope results in a larger numerical value.

Conditions: Both graphs are evaluated at the same input $x=4$.; Both graphs are curving upward near $x=4$.

Although both graphs curve upward (indicating a positive second derivative), the narrow parabola has a much more rapid increase in slope around $x=4$ compared to the wider parabola. Since the second derivative measures the rate of change of the slope, a faster change in slope results in a larger numerical value.

Conditions: Both graphs are evaluated at the same input $x=4$.; Both graphs are curving upward near $x=4$.

A finite Riemann sum uses $\approx$ because it approximates the volume under a curved surface $f(x,y)$ using flat-topped boxes, which inevitably introduces local errors. Exact equality holds only in the limit as the partition norm $\|P\| \to 0$, where the approximation converges to the definite integral.

Conditions: The sum involves a finite number $n$ of subrectangles.; The surface $f(x,y)$ is curved (not a flat plane).

A finite Riemann sum uses $\approx$ because it approximates the volume under a curved surface $f(x,y)$ using flat-topped boxes, which inevitably introduces local errors. Exact equality holds only in the limit as the partition norm $\|P\| \to 0$, where the approximation converges to the definite integral.

Conditions: The sum involves a finite number $n$ of subrectangles.; The surface $f(x,y)$ is curved (not a flat plane).

The coefficient formula for a Maclaurin series is $c_n = f^{(n)}(0) / n$!. This means the nth coefficient of the power series is found by taking the nth derivative of the function, evaluating it at $x=0$, and dividing by n factorial.

Conditions: The series is expanded around $x=0$.; The function must have derivatives up to order n at 0 for the displayed formula to make sense.; The function is assumed to have a power series representation.

The coefficient formula for a Maclaurin series is $c_n = f^{(n)}(0) / n$!. This means the nth coefficient of the power series is found by taking the nth derivative of the function, evaluating it at $x=0$, and dividing by n factorial.

Conditions: The series is expanded around $x=0$.; The function must have derivatives up to order n at 0 for the displayed formula to make sense.; The function is assumed to have a power series representation.

To evaluate the right-hand limit ($\lim_{x \to c^+} f(x)$), trace the graph starting from x-values greater than c and moving leftward toward c. Observe the height (y-value) that the curve approaches.

Conditions: Evaluating the limit from the right side (inputs strictly greater than c).; The graph shows a clear trend toward a finite y-level.

To evaluate the right-hand limit ($\lim_{x \to c^+} f(x)$), trace the graph starting from x-values greater than c and moving leftward toward c. Observe the height (y-value) that the curve approaches.

Conditions: Evaluating the limit from the right side (inputs strictly greater than c).; The graph shows a clear trend toward a finite y-level.

Because the recursive formula needs a previous forecast, the video demonstrates a simple initialization method: set the first forecast equal to the first actual observation. In the example this gives $F(1)=A(1)=1103$.

Conditions: Applies when no earlier forecast is available.; Shown specifically for Week 1 in the example.

Because the recursive formula needs a previous forecast, the video demonstrates a simple initialization method: set the first forecast equal to the first actual observation. In the example this gives $F(1)=A(1)=1103$.

Conditions: Applies when no earlier forecast is available.; Shown specifically for Week 1 in the example.

The Taylor series for $e^x$ centered at 0 simplifies to sum $x^n/n$! because every derivative of $e^x$ is exactly $e^x$. When evaluating the nth derivative at the center $x=0$, the result is always $e^0$, which equals 1.

Conditions: $f(x)=e^x$; center $a=0$

The Taylor series for $e^x$ centered at 0 simplifies to sum $x^n/n$! because every derivative of $e^x$ is exactly $e^x$. When evaluating the nth derivative at the center $x=0$, the result is always $e^0$, which equals 1.

Conditions: $f(x)=e^x$; center $a=0$

Continuity ensures that as the interval width $h$ approaches zero, the average value of $f$ over $[x, x+h]$ converges to the instantaneous value $f(x)$. Without continuity, the local behavior might oscillate wildly or have jumps, preventing the limit of the difference quotient from settling on a single well-defined value $f(x)$.

Conditions: $A(x) = \int_a^x f(t) dt$; $f$ is continuous at $x$

Continuity ensures that as the interval width $h$ approaches zero, the average value of $f$ over $[x, x+h]$ converges to the instantaneous value $f(x)$. Without continuity, the local behavior might oscillate wildly or have jumps, preventing the limit of the difference quotient from settling on a single well-defined value $f(x)$.

Conditions: $A(x) = \int_a^x f(t) dt$; $f$ is continuous at $x$

The volume under a surface $z=f(x,y)$ over a rectangular region $[a,b] \times [c,d]$ is defined through a four-step limiting process using double Riemann sums. First, the region is partitioned into subrectangles with area $\Delta A_k = \Delta x_k \Delta y_k$, and a sample point $(x_k,y_k)$ is chosen in each. The volume is approximated by the sum $\sum_{k=1}^n f(x_k,y_k) \Delta x_k \Delta y_k$, and the exact volume is defined as the limit of this sum as the partition norm $\|P\| \to 0$.

Conditions: The base region must be a rectangle $[a,b] \times [c,d]$.; The function $f(x,y)$ defines the height of the surface.; The limit is taken as the largest subrectangle area approaches zero.

The volume under a surface $z=f(x,y)$ over a rectangular region $[a,b] \times [c,d]$ is defined through a four-step limiting process using double Riemann sums. First, the region is partitioned into subrectangles with area $\Delta A_k = \Delta x_k \Delta y_k$, and a sample point $(x_k,y_k)$ is chosen in each. The volume is approximated by the sum $\sum_{k=1}^n f(x_k,y_k) \Delta x_k \Delta y_k$, and the exact volume is defined as the limit of this sum as the partition norm $\|P\| \to 0$.

Conditions: The base region must be a rectangle $[a,b] \times [c,d]$.; The function $f(x,y)$ defines the height of the surface.; The limit is taken as the largest subrectangle area approaches zero.

To draw the graph of a continuous uniform distribution, plot the lower bound 'a' and the upper bound 'b' on the horizontal x-axis. Then, plot the constant probability density value $\frac{1}{b-a}$ on the vertical y-axis.

Conditions: You know the values of the lower bound a and the upper bound b.; You have calculated the constant probability density value $\frac{1}{b-a}$.

To draw the graph of a continuous uniform distribution, plot the lower bound 'a' and the upper bound 'b' on the horizontal x-axis. Then, plot the constant probability density value $\frac{1}{b-a}$ on the vertical y-axis.

Conditions: You know the values of the lower bound a and the upper bound b.; You have calculated the constant probability density value $\frac{1}{b-a}$.

In this forecasting method, $\alpha$ is the smoothing constant that determines the weight assigned to the most recent actual demand. The term $(1-\alpha)$ is the complementary weight applied to the previous forecast.

Conditions: $\alpha$ is fixed at 0.4 in the worked example.; $(1-\alpha)$ is calculated as 0.6.; The weights must sum to 1 for the formula to represent a valid weighted average.

In this forecasting method, $\alpha$ is the smoothing constant that determines the weight assigned to the most recent actual demand. The term $(1-\alpha)$ is the complementary weight applied to the previous forecast.

Conditions: $\alpha$ is fixed at 0.4 in the worked example.; $(1-\alpha)$ is calculated as 0.6.; The weights must sum to 1 for the formula to represent a valid weighted average.

To find the left-hand limit as x approaches a value c (denoted $\lim_{x \to c^-} f(x)$), observe the behavior of the function's curve strictly for input values less than c, moving towards c. If the y-values grow without bound in either the positive or negative direction, the finite real limit does not exist.

Conditions: The function has a vertical asymptote or discontinuity at $x=c$.; You are evaluating the limit from the left side (inputs strictly less than c).

To find the left-hand limit as x approaches a value c (denoted $\lim_{x \to c^-} f(x)$), observe the behavior of the function's curve strictly for input values less than c, moving towards c. If the y-values grow without bound in either the positive or negative direction, the finite real limit does not exist.

Conditions: The function has a vertical asymptote or discontinuity at $x=c$.; You are evaluating the limit from the left side (inputs strictly less than c).

After substituting $t=0$ and $F(\omega) = 2 \frac{\sin \omega}{\omega}$ into the inverse Fourier transform formula, the equation becomes $1 = \frac{1}{2\pi} \int_{-\infty}^{\infty} 2 \frac{\sin \omega}{\omega} d\omega$. The constant 2 in the numerator of the integrand cancels with the constant 2 in the denominator of the prefactor $\frac{1}{2\pi}$.

Conditions: The inverse Fourier transform formula is used: $f(0) = \frac{1}{2\pi} \int_{-\infty}^{\infty} F(\omega) d\omega$.; $f(0) = 1$.; $F(\omega) = 2 \frac{\sin \omega}{\omega}$.

After substituting $t=0$ and $F(\omega) = 2 \frac{\sin \omega}{\omega}$ into the inverse Fourier transform formula, the equation becomes $1 = \frac{1}{2\pi} \int_{-\infty}^{\infty} 2 \frac{\sin \omega}{\omega} d\omega$. The constant 2 in the numerator of the integrand cancels with the constant 2 in the denominator of the prefactor $\frac{1}{2\pi}$.

Conditions: The inverse Fourier transform formula is used: $f(0) = \frac{1}{2\pi} \int_{-\infty}^{\infty} F(\omega) d\omega$.; $f(0) = 1$.; $F(\omega) = 2 \frac{\sin \omega}{\omega}$.