Justify $\Theta(n^7 \log n)$ for $g(n)$

  • Context: MHB 
  • Thread starter Thread starter evinda
  • Start date Start date
Click For Summary
SUMMARY

The asymptotic complexity of the function $g(n) = 8n^7 \log n + 18 \log{13} + 16 \log n^n + 12 n^{\frac{5}{2}}$ is definitively $\Theta(n^7 \log n)$. The dominant term is $n^7 \log n$, and it can be justified by demonstrating that all other terms are asymptotically smaller than this dominant term for sufficiently large $n$. Specifically, for $n \geq 2$, it is shown that $18 \log{13}$, $16n \log n$, and $12n^{\frac{5}{2}}$ are all less than or equal to $n^7 \log n$, confirming that $g(n)$ satisfies the definition of $\Theta(n^7 \log n)$.

PREREQUISITES
  • Understanding of asymptotic notation, specifically $\Theta$ notation.
  • Familiarity with logarithmic and polynomial functions.
  • Basic knowledge of limits and growth rates of functions.
  • Ability to manipulate inequalities involving functions of $n$.
NEXT STEPS
  • Study the formal definition of $\Theta$ notation and its implications in algorithm analysis.
  • Learn about comparing growth rates of different functions, particularly polynomials and logarithmic functions.
  • Explore examples of asymptotic analysis in algorithm complexity, focusing on dominant terms.
  • Investigate the use of limits in proving asymptotic relationships between functions.
USEFUL FOR

Students and professionals in computer science, particularly those focused on algorithm analysis, complexity theory, and mathematical foundations of computer science.

evinda
Gold Member
MHB
Messages
3,741
Reaction score
0
Hi! (Smile)

I want to find the asymptotic complexity ($\Theta$) of the function $g(n)=8n^7 \log n+18 \log{13}+16 \log n^n+12 n^{\frac{5}{2}}$.

We see that the dominant term is $n^7 \log n$, but how could we justify it? (Thinking)

Does this mean that : $g(n)=\Theta(n^7 \log n)$ ? :confused:
 
Technology news on Phys.org
According to definition we have to find constants such that

$$c_1 f(n) \leq g(n) \leq c_2 f(n) $$

is held true for all $n\geq N_0$ , Can you find such constants ?
 
ZaidAlyafey said:
According to definition we have to find constants such that

$$c_1 f(n) \leq g(n) \leq c_2 f(n) $$

is held true for all $n\geq N_0$ , Can you find such constants ?

Yes, but.. how can I explain that $n^7 \log n$ is the dominant term? (Thinking)
 
You can try to verify it by definition, that is,
$$g(n) = \Theta(n^7 \log n) \Leftrightarrow \exists c>0, \exists n_0 \in \mathbb{N}, \forall n \geq n_0: 0 \leq g(n) \leq c n^7 \log n$$

I would do the following: try to show that all terms (with exception of the first term obviously) of $g(n)$ are smaller than $n^7 \log n$ for a certain $n_0 \in \mathbb{N}$.

First, note that $\forall n \geq 2: 18 \log 13 \leq n^7 \log n$. This follows from the fact that $\log n$ and $n^7$ are increasing functions forall $n \in \mathbb{N}_0$.

Second, we can write $16 \log n^n = 16n \log n$. We want to search for which values of $n \in \mathbb{N}_0$ we have: $16 n \log n \leq n^7 \log n$ or equivalently $16 n \leq n^7$. Since $n^7$ increases much faster than $16n$ for the natural numbers (without zero) we can conlude that $\forall n \geq 2: 16n \leq n^7$, therefore $ \forall n \geq 2: 16n \log n \leq n^7 \log n$.

For the last term we can use a similar argument to show that $\forall n \geq 2: 12n^{\frac{5}{2}} \leq n^7 \log n$.

At this point we have shown that all terms are smaller than $n^7 \log n$.
What's the conclusion now?
 
At the beginning, before I verify that $g(n)=\Theta(n^7 \log n)$ can I just say the following?

We see that the dominant term of the function $g(n)$ is $n^7 \log n$, so: $g(n)=\Theta(n^7 \log n)$ ?

Or is there a better way to explain why we suppose that $g(n)=\Theta(n^7 \log n)$? (Thinking)
 
By definition

$$T(n) = \Theta( G(n) ) \, \implies \,\, \lim \frac{T(n)}{G(n)} \to c \, ; \,\, 0<c<\infty $$

Suppose that $f(n) = a(n) +g(n)$ where $a(n)$ is the dominant term so

$$\lim_{n \to \infty } \frac{a(n)+g(n)}{a(n)} = 1+ 0 = 1 $$

So $T(n) = \Theta (a(n)) \blacksquare $.
 

Similar threads

  • · Replies 8 ·
Replies
8
Views
3K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 24 ·
Replies
24
Views
4K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 1 ·
Replies
1
Views
1K
  • · Replies 15 ·
Replies
15
Views
8K
  • · Replies 7 ·
Replies
7
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
  • · Replies 17 ·
Replies
17
Views
4K
  • · Replies 6 ·
Replies
6
Views
2K