WHAT IS THE DIFFERENT algorithm of advantage and amp disadvantage?

Answer 1

Different algorithms do different things, so it makes no sense to compare them. For example, the accumulate algorithm is an algorithm which performs the same operation upon every element of a container, whereas a sorting algorithm sorts the elements of a container. Each specific algorithm requires a different set of concepts. An accumulate algorithm requires a data sequence with at least forward iteration and elements which support the operation to be performed, whereas a sorting algorithm generally requires random access iterators and elements that support a given comparison operation (such as the less-than operator).

Even if two algorithms have the exact same time and space complexities, it does not follow that both will complete the task in the same time. For instance, the accumulate algorithm is a linear algorithm with a time-complexity of O(n) regardless of which operation is being performed. However, the complexity of the operation itself can greatly affect the actual time taken, even when the operations have exactly the same time-complexity. For instance, if we use the accumulate algorithm in its default form (to sum all the elements in a data sequence), the operation itself has a constant-time complexity of O(1). If we choose another operation, such as scaling each element and summing their products, it will take much longer to complete the algorithm (possibly twice as long) even though the operation itself has the exact same time-complexity, O(1).

Consider the time-complexity of adding one value to another:

a += b

This has to be a constant-time operation because the actual values of a and b have no effect upon the time taken to produce a result in a. 0 += 0 takes exactly the same number of CPU cycles as 42 += 1000000.

Now consider the operation to scale and sum:

a += b * 42

Here, 42 is the scalar. This also has to be a constant-time operation, but it will take longer to physically perform this operation compared to the previous one because there are more individual operations being performed (roughly twice as many).

The only way to compare algorithms is to compare those that achieve exactly the same goal but do so in different ways. Only then does comparing their respective time-complexity make any sense. Even so, time-complexity is merely an indication of performance so two sorting algorithms with the exact same time-complexity could have very different runtime performance (it depends on the number and type of operations being performed upon each iteration of the algorithm). Only real-world performance testing can actually determine which algorithm gives the best performance on average.

With sorting algorithms, we often find one algorithm ideally suited to sorting small sequences (such as heap sort) and others ideally suited to larger sets (such as merge sort). Combining the two to create a hybrid algorithm would give us the best of both worlds.