Bulletin of Electrical Engineering and Informatics

Received Dec 1, 2022 Revised May 28, 2023 Accepted Jun 5, 2023 In a competitive environment, the ability to rapidly and successfully scale up new business models is critical. However, research shows that many new business models fail. This research looks at hybrid methods for minimizing constraints and maximizing opportunities in large data sets by examining the multivariable that arise in user behavior. E-metric data is being used as assessment material. The analytical hierarchy process (AHP) is used in the multi-criteria decision making (MCDM) approach to identify problems, compile references, evaluate alternatives, and determine the best alternative. The multi-objectives genetic algorithm (MOGA) role analyzes and predicts data. The method is being implemented to expand the information base of the strategic planning process. This research examines business sustainability along two critical dimensions. First, consider the importance of economic, environmental, and social evaluation metrics. Second, the difficulty of gathering information will be used as a predictor for making long-term business decisions. The results show that by incorporating the complexity features of input optimization, uncertainty optimization, and output value optimization, the complexity prediction model (MPK) achieves an accuracy of 89%. So that it can be used to forecast future business needs by taking into account aspects of change and adaptive behavior toward the economy, environment, and social factors.


INTRODUCTION
Increasing the productivity of business processes has become a major issue, both in academia and in business, because organizations must provide effective and efficient results [1], [2].The impact of commercial digital business actors' characteristic patterns varies greatly and is highly competitive for users with diverse desires [3].This is demonstrated by the continued expansion of business actors by prioritizing profit-oriented, revenue-generating activities, as well as consideration in increasing the competitiveness of the number of merchants obtained electronically, thereby providing users with a plethora of options for facilitating their transactions [4], [5].
There must be anticipation for business actors by preventing disruptions in order for the business to be sustainable, specifically by providing a significant social impact caused by business actors and users [6], [7].The current issue is that people have a variety of digital shopping options that business actors must consider; as a result, business metrics are used in determining investment in long-term business opportunities [8], [9].Multi criteria decision making (MCDM) is a decision making technique that chooses the  ISSN: 2302-9285 Bulletin of Electr Eng & Inf, Vol. 12, No. 6, December 2023: 3697-3705 3698 best option among several alternatives based on specific criteria.Criteria are commonly used decision making measures, rules, or standards [10]- [12].
One of the MCDM methods is the analytical hierarchy process (AHP), which works to decompose complex multi factor or multi criteria problems into a hierarchy [13].The first level represents the goal, and the subsequent levels represent factors, criteria, and sub criteria [14], [15].To select individuals with the same rank, a split mechanism is used.It plans to investigate several business process model variants that will be processed using the AHP.As a basis for decision making, such as alternatives, attributes, conflicts between criteria, and decision weights [16], [17].Multi-objectives genetic algorithm (MOGA) is a multi-objective optimization algorithm that, by allowing the user to explore different regions of the solution space, is particularly well suited to solving multi-objective optimization problems [18], [19].As a result, a more diverse set of solutions can be sought, with more variables that can be optimized concurrently [20], [21].

METHOD
The first step is to collect data from e-metrics.Furthermore, the data is processed using the AHP method, namely special weighting by determining alternatives, criteria, and attributes.The examination results were re-tested for each alternative using the MOGA combinatorial algorithm approach.MOGA is used to find populations through selection, crossover, and mutation.The research steps can be seen in Figure 1.

Multi-criteria decision making
MCDM is a systematic procedure for changing complex problem decisions with a predefined sequence of steps that can be followed and used to aid in rational decision making [22]: i) create a model that describes a structured system, including component relationships and interaction criteria; ii) goals must be established; iii) develop relevant criteria for differentiating between desirable and undesirable goals; iv) construct and evaluate potential alternatives; v) experiment with various options to see if they meet your objectives; vi) consider the consequences of the alternatives; and vii) weighing and sorting alternative options based on the decision maker's preferences.

Concept analytic hierarchy process
AHP algorithms provide structured and systematic work assignments for problem solving, particularly in situations when there are several criteria and alternatives to consider.This aids in the completion of subordinate tasks, allows for the comparison of monetary factors, and provides a clear process for evaluating and prioritizing priorities.The AHP method has a hierarchical structure that divides complicated decision issues into smaller pairwise comparisons.The AHP algorithm is explained in detail in Figure 2.

Figure 2. Hierarchy process AHP
A matrix is used to implement the AHP model [23].For example, suppose an operating subsystem has n elements, i.e. operating elements C1, C2,..., Cn.The comparisons shown in the Table 1 are the results of pairwise comparisons of the operating elements [24], [25].Matrix ( × ) is a reciprocal matrix with  elements i.e.,  1 ,  2 , … ,   to be compared.As shown in (1), the results of the pairwise comparisons between (  ,   ) can be represented as a matrix [26]: A comparison matrix with intensity values represents pairwise comparisons of items using numerical intensity values.This intensity number represents the importance or preference given to one aspect above another as in Table 2 Once a comparison matrix containing intensity values is created, it may be used as the foundation for additional calculations such as normalization to generate priority weights for components and make educated judgments based on their relative relevance or preference.Score   ,   with ,  = 1, 2, … ,  obtained from a higher value.The eigen value of A is n when this matrix is multiplied by the column vector  =  1 ,  2 , … ,   [27].
In general, the variable n in the illustration can be substituted by a vector  in [28]: Every   satisfying in (3) is referred to as an eigen value, and every vector  satisfying the above equation is referred to as an eigen vector [29], [30].The comparison matrix is said to be consistent in this case if the consistency index () value is less than. parameter deviation   from  is formulated as (4):

Multi-objective genetic algorithm
MOGA assesses individual survival of the fittest during executive generation.Each generation is made up of a population of character strings that function similarly to chromosomes [31], [32].In general, the purpose can be defined as (5): Where  / explains why functions, input variables, model coefficients, and constant coefficients should be maximized or minimized     [33 ], [34] .

RESULTS AND DISCUSSION
The first step is to create an objective hierarchy of criteria and to identify alternative criteria.The second step is to compute the criteria and compare the alternatives.To begin, the criteria in hierarchical construction must be defined at three levels: top, middle, and bottom.The upper level defines the objectives, the middle level defines the criteria and sub criteria, and the lower level defines the lower alternatives.The third step is to rank the criteria, sub criteria, and alternatives in order of importance.

Criteria weighting measures analytical hierarchy process
A criterion weighting measure is used in the decision-making process to evaluate the relative value or priority of distinct criteria.Based on pairwise comparisons, AHP provides a structured approach for assigning priority weights to criteria.The criteria weighing steps in AHP involve the following steps: a. Create a pairwise comparison matrix between weight criteria Pairwise comparison: determining the relative importance or preference of two criteria.Each criterion is compared to the others, and an intensity score or rating is assigned to reflect its relative relevance based on (1).The elements compared in the matrix are shown in Table 3.The comparison value of each criterion element is shown in Table 4. b.Calculating the priority weight Table 5 shows the calculations used to obtain the priority vector or total priority value (TPV).Priority vector or TPV refers to the set of priority weights assigned to elements in a decision hierarchy.These weights represent the relative importance of each element in the decision objectives shown in Table 6.Calculate the value of the consistency ratio from the value of random 6 elements then choose 2.3 then the value of CR is: The criterion matrix's consistency ratio is 0.08 (0.08 0.1), showing good consistency, because a value of 0.08 implies that the consistency ratio of the comparative research outcomes is 8%.Because it is less than 0.1 (10%), the value is acceptable.

Stages multi-objectives genetic algorithm
The MOGA algorithm continues iterating until the termination condition is met.It aims to converge toward a set of solutions representing trade-offs or Pareto fronts between conflicting goals.The final output of the MOGA algorithm is a set of non-dominant solutions that represent the best trade-off solutions for multiobjective optimization problems in (5).Table 7 shows the lower and upper threshold values as the basis for finding versatile values for various scenarios.In multi-objective optimization, fitness is determined by evaluating each individual's performance across multiple objective functions.The objective function is usually defined based on the specific problem being addressed.
In the Table 8 several scenarios were carried out to evaluate each individual in several objective functions.The MOGA approach solves the majority of the issues.MOGA generates more decision information based on the AHP process, as shown in Figure 3, based on the scenarios generated in Tables 2 and 4. The MOGA algorithm performs convergence behavior analysis by tracking the development of Pareto front or nondominance solutions from element to element.The convergence sign in Figure 3 The optimization measure is calculated using the optimization method over the predicted range each time the e-metric data is sampled in this strategy [35].The complexity prediction model (MPK) optimization model is shown (6): Modeling is generated based on input complexity (R), uncertainty (A), and output value (D).Therefore, it means:  = 0.     = 0.                   ×    ×    ×    × The derivative of the prediction model and the period complexity of the future state are as (8): Analysis of user behavior on e-metric data is done by looking at user habits in business metrics.The study that has been carried out is based on the estimated maximum limit based on the prediction results in Table 9. Figure 4 shows the results of the optimization model.Where the MPK approach optimizes 3 parts of e-metric data.Figure 4(a) optimizes the input, Figure 4(b) optimizes the uncertainty, and Figure 4(c) optimizes the output value.It can be concluded that the approach provides information about strategies and alternatives.The MCDM-MOGA hybrid model was developed into a MPK based on (6) to (8), which responds to predictions of future user behavior by producing the highest value accuracy of 89% at the output value optimization based on complexity results prediction

CONCLUSION
In this study, a complexity prediction model was developed to address the challenges posed by multiobject complexity on business uncertainty problems.This model aims to provide a comprehensive understanding of the factors contributing to complexity and offer insights for managing a dynamic business environment.research shows the importance of considering multiple objectives and the impact of uncertainty on complexity.The developed prediction model incorporates various factors such as organizational structure, resource allocation, technological advances, market dynamics, and external factors to assess and predict the level of complexity.The results show that the complexity prediction model is promising in capturing and predicting patterns of complexity, serving as a tool for organizations to assess the level of complexity, identify critical factors, and make decisions to optimize performance and reduce potential risks, where MCDM describes data by looking at opportunities in making decisions based on many criteria, the MOGA scenario looks for customer habits towards business metrics, resulting in an MPK method with an accuracy of 89% and can be used in predicting future business needs.

Figure 1 .
Figure 1.State of the art research Bulletin of Electr Eng & Inf ISSN: 2302-9285  Complexity prediction model: a model for multi-object complexity in consideration … (Rahmad B. Y. Syah) 3699

Figure 3 .
Figure 3. Predicted output; (a) behavior of each element is greater and (b) behavior of each element is smaller

Table 2 .
Comparison matrix with intensity value

Table 3 .
Initial criteria element

Table 4 .
Criteria comparison matrix

Table 6
Next, compute the  by adding   minus  (number of criteria) and then subtracting  minus 1 as show in (4).

Table 7 .
Lower and upper bounds for the optimization problem