Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

Investigating the Composition of Apriori Algorithm and

Metaheuristic Algorithms (Genetic and PSO)

Mahmoud Alimoradi

The APRIORI method is a basic method for finding frequent items. Basic methods are used to solve
basic problems in computer science. Computer science issues have to be investigated from a variety of
perspectives and problems encountered in the field of computer science. In order of priority, we seek to
remove restrictions and speed up the resolution of problems.

To explore items, we need to find those items that are repeatedly and at a minimum. Also, in this way,
we must leave the data that is not worth the time to reduce the system load. The basic method
APRIORI is also one of the solutions introduced to the world of computers for the list of itemized
items.

The basis of this approach is that every single duplicate item that has been seen should be discarded.
These low-repeat items also give rise to non-repetitive items in the extension of the subset. Well, it is so
good here. We have items We pass slowly and in vain, and eventually we reach a valuable and frequent
item set. That is also the goal.

But the first problem that can be seen clearly is that we have to examine the entire setup and pay for it
in a huge amount of data. What is in terms of time or system overhead, which is far from optimal. Is.
To date, various approaches have been proposed that each of these methods has the advantage that
alongside the advantages have a disadvantage and are usually such that they are optimal in terms of
time, but when we look at their result, waste Sometimes we see frequent data.
It means that, at a reasonable time, we give repeated questions, but do not fully disclose repeated data.
Of course, in some cases, it is necessary to ignore these cases in computer science.
We combine the basic APRIORI method with Genetic Algorithm and PSO in a new way, which, in
addition to optimizing the important parameter of time, minimizes the number of missing items to the
minimum.

We call these new methods GA-APRIORI and PSO-APRIORI. These methods, in addition to
minimizing the time, as well as other methods, are efficient in terms of efficiency. This is the solution
to this problem.

In our genetic method, with the specific algorithm outlined below, we define a current population. This
current population is filled in with the items in length 1 in the first direction. Further, if these items
should be repeated frequently to the population of repetitive or otherwise eliminated In the sequel, we
have the genetic algorithm We have the parent 1 and 2 that use the genetic algorithm to create new
items and, with APRIORI, remove the non-repetitive data from the cycle. This process continues until
the end of production by GA. Find out
1: Input: T: Transactional database. MinSup: MinimumSupport user's threshold.
2: Output: F: The set offrequent Itemsets.
3: F ← Ø.
4: CurrentPopulation ← FindFrequentOneItemset ().
5: while CurrentPopulation ~ϵ Ø do
6: F ← F CurrentPopulation.
7: NewPopulation ← Ø.
8: for each of two individuals (parent1, parent2) ϵ CurrentPopulation do
9: NewPopulation ← NewPopulation U Crossover (Parent1, Parent2).
10: end for
11: for each infrequent indiv NewPopulation do
12: while infrequent individua
13: mutation (indiv)
14: end while
15: end for
16: F ← F U NewPopulation.
17: CurrentPopulation ← Selection (NewPopulation).
18: end while
19: return F

But in the PSO algorithm, we need to get inspired by this algorithm and combine it with APRIORI to
our new innovative algorithm. The following algorithm is inspired by the movement of birds to reach
the food.

Consider the category of birds that are looking for food. All birds at one time are going to the only
source of food. Imagine them in several categories. Each bunch has a distance from the source. Each
bird is more than the other Both batches have a source limit, and this is the same as for all birds. Now,
it should be kept in mind that each bird is flying at a fast pace. This is the basis of this algorithm. Now
Imagine our data particles as birds.

As you see in the algorithm, we define the particles, and, according to the algorithm, we connect them
to the best of the region and the whole, and then combine them together. If it was repeated, the frequent
set It joins with the APRIORI algorithm, otherwise it is separated from the cycle. In the end, we also
give us a bunch of frequent particles of these particles as output.

1: Input: T: Transactional database. MinSup: MinimumSupport user's threshold.


2: Output: F: The set offrequent Itemsets.
3: F ← Ø.
4: Particles ← FindFrequentOneItemset ().
5: F ← F U Particles.x.
6: S ← Particles.x.
7: while S ~ϵ Ø do
8: S ← Ø.
9: for each particle pi do
10: Join pi.vi ← (BestGlobal, BestLocali)
11: pi.xi ← Combining (pi.xi, pi.vi).
12: if pi.xi is frequent then
13: S ← U pi.xi.
14: end if
15: end for
16: F ← F U S.
17: end while
18: return F

According to the tests, these two methods, in terms of both time and optimality, have surpassed
comparisons of the other competitors compared to the previous ones. But, as it seems, this way is to
provide better and better algorithms Still open yet.
It may be possible to combine the algorithms of meta-innovative algorithms with basic methods into
better and more dynamic algorithms, which would be the starting point for them.

References
[1] Agarwal, R., & Srikant, R. (1994, September). Fast algorithms for mining association rules. In
Proc. of the 20th VLDB Conference (pp. 487-499).
[2] Srinivas, M., & Patnaik, L. M. (1994). Genetic algorithms: A survey. computer, 27(6), 17-26.
[3] Eberhart, R., & Kennedy, J. (1995, October). A new optimizer using particle swarm theory. In
Micro Machine and Human Science, 1995. MHS'95., Proceedings of the Sixth International
Symposium on (pp. 39-43). IEEE.
[4] Djenouri, Y., & Comuzzi, M. (2017). Combining Apriori heuristic and bio-inspired algorithms
for solving the frequent itemsets mining problem. Information Sciences, 420, 1-15.
[5] Azgomi, H., & Mahjur, A. (2013). A Solution for Calculating the False Positive and False
Negative in LSH Method to Find Similar Documents.Journal of Basic and Applied Research,
3(1s), 466-472.
[6] Azgomi, H. (2013). Finding similar documents by removing redundant characters.(master
dissertation). Islamic Azad University of Shabestar, Shabestar, Iran.
[7] Azgomi, H., Ghasemi Mahsayeh, M., & Mohammadpour, M. (2013). A Method for Finding
Similar Documents Based on Selection a Range for Substring Lenght in Shingling Method.Reef
Resources Assessment and Management Technical Paper, 38(5).
[8] Azgomi, H., Ghasemi Mahsayeh, M., Mohammadi, M., & Moradi, M. (2014).A Method for
Finding Similar Documents Relying on Adding Repetition of Symbols in Length Based
Filtering. Indian Journal of Scientific Research, 2(1), 81-84.
[9] Azgomi, H., & Ghasemi Mahsayeh, M (2014). Finding Similar Items In Data mining: Shingling
Method, Presentation, Iran.
[10] Azgomi, H., & Ghasemi Mahsayeh, Ma. (2013). Finding Similar Documents by
Repetition of Symbols in Length Based Filtering. International Conference on Computer,
Information Technology and Digital Media, Tehran, Iran.
[11] Azgomi, H. (2015). Apriori Source Code in VB, Java and C#.
[12] Ghasemi Mahsayeh, M., & Mahjur, A. (2013). A Method for Finding Similar Documents
on the Basis of Repetition-Based Filtering.Journal of Basic and Applied Research, 3(1s), 603-
607.
[13] Ghasemi Mahsayeh, M. (2013). Finding similar documents relying on repetition based
filtering.(master dissertation). Islamic Azad University of Shabestar, Shabestar, Iran.
[14] Sohrabi, M. K., & Azgomi, H. (2017). TSGV: a table-like structure-based greedy
method for materialized view selection in data warehouses. Turkish Journal of Electrical
Engineering & Computer Sciences, 25(4), 3175-3187.
[15] Sohrabi, M. K., & Azgomi, H. (2017). Parallel set similarity join on big data based on
Locality-Sensitive Hashing. Science of Computer Programming, 145, 1-12.
[16] Azgomi, H., & Sohrabi, M. K. (2018). A game theory based framework for materialized
view selection in data warehouses. Engineering Applications of Artificial Intelligence, 71, 125-
137.

[17] Azgomi, H., & Sohrabi, M. K. (2018). Finding similar documents using frequent pattern
mining methods. International Journal of Uncertainty, Fuzziness and Knowledge-Based
Systems.
[18] Azgomi, H. (2018). Thesis: A game theory based framework for materialized view
selection in data warehouse. (Doctoral dissertation). Islamic Azad University of Semnan,
Senman, Iran. 10.13140/RG.2.2.20178.45766.
[19] Azgomi, H., & Sohrabi, M. K. (2017). RTLTDS Dataset. 10.13140/RG.2.2.35186.81608.
[20] Azgomi, H. (2017). RTLTDS Data Warehouse. 10.13140/RG.2.2.35513.88166.

[21] Sohrabi, M. K., & Azgomi, H. (2018). Evolutionary game theory approach to
materialized view selection in data warehouses. Knowledge-Based Systems.
[22] Sohrabi, M. K., & Azgomi, H. A Survey on the Combined Use of Optimization Methods
and Game Theory. Archives of Computational Methods in Engineering, 1-22.

You might also like