Comparison Of Population Prediction Of Bengkulu Selatan Regency With Python And SPSS

: Bengkulu Selatan Regency is one of the oldest regencies in Bengkulu Province, with an area of 1,186.09 km² consisting of 11 sub-districts, 142 villages, and 16 urban-villages. Based on aggregate population data from the Population and Civil Registration Service of Bengkulu Selatan Regency, in 2023, the population was 174,936 people, consisting of 88,749 men and 86,142 women, with a population growth of 1,810 people. The population aspect is an important variable in determining the carrying capacity and capacity related to development and regional planning. In this study, researchers compared the population predictions of Bengkulu Selatan Regency using Python and the SPSS application. Population predictions use multiple linear regression analysis and are packaged in a web-based application using Python. This study uses the variables of birth, death, in-migration, and out-migration as independent variables and population as the dependent variable.The prediction results using Python and SPSS applications produce the same linear regression equation model to predict population size, namely 𝑌 = 160036.652 −1.346 births + 2.048 deaths + 6.371 in-migration + 1.618 out-migration. The coefficient of determination is 0.982, indicating that the in-migration variable has a significant positive effect, while the other three variables do not have a significant effect.


Introduction
South Bengkulu Regency is one of the oldest regencies in Bengkulu Province which was formed in 1949 since the division of Bengkulu Province from South Sumatra Province, currently South Bengkulu Regency is led by Gusnan Mulyadi, SE as the Regent of South Bengkulu (Mildawani et al., 2022).Based on aggregate population data from the Population and Civil Registry Office of South Bengkulu Regency in 2023, the population of South Bengkulu Regency is 174,936 people consisting of 88,749 men and 86,142 women with a population growth of 1,810 people or 0.29%.The amount of population growth can affect regional development planning so that appropriate analysis is needed (Spooner et al., 2020).
The regional medium-term development plan is carried out every 5 (five) years and evaluated every 1 (one) year.The evaluation of the development plan is based on sociopopulation analysis, in this case the population.Population surges can cause problems in https://penerbitadm.pubmedia.id/index.php/KOMITEKthe future if not handled properly (Schafer et al., 2021).The population also affects the capacity and carrying capacity of an area.Thus, the p o p u l a t i o n data can be used as the main reference in considering the preparation of population projections that can provide an overview to the Regional Government in preparing the next planned and quality development plan (Mostafavi et al., 2020).
The results of previous research as support for this study include (Prawidana et al 2022) stating that the results of analysis and discussion of data mining using multiple linear regression methods about (Park et al., 2019).
The prediction of population growth rate in South Jakarta has an accurate calculation level as indicated by the Root Mean Square Error value of 0.43 < 1.0.According to (Widia et al, 2022) the results of estimating the amount of population growth using multiple linear regression methods are the right method for predicting population with a high level of accuracy and hope that the results of this study can be input to the Gunung Malela Sub-district Office in anticipating population growth rates (Boo & Choi, 2022).In a study conducted by Widia, et al, there was an increase in population in Gunung Malela District of 496 people throughout 2021-2025. (Candra et al, 2023) ) state that the multiple linear regression method can provide an estimate of population increase up to an accuracy level of 94.5%.(Ayu Wulandari, et al 2022) stated that the multiple linear regression method can provide population estimation results with a high level of accuracy (Abdelaal et al., 2019), and suggested using the multiple linear regression method to be further used in calculations that make it easier for BPS to estimate the population rate (Islam et al., 2020).(Purwadi, et al 2019) stated that the analysis results obtained from data mining using multiple linear regression methods regarding population rate estimation can assist BPS in knowing what criteria can affect population growth rates and the multiple linear regression method can be implemented in predicting population growth rates with fairly accurate results (Rimal et al., 2019).
Based on previous research that has been presented, the researcher took the research title on the comparison of population prediction South Bengkulu Regency with python and SPSS applications.The title was taken as a form of continuation of previous research (Rasjid et al., 2021).This research will implement multiple linear regression analysis packaged in an application-based (Kammar-García et al., 2021).

Problem Analysis
Researchers identified problems related to the comparison of population prediction in South Bengkulu Regency using python and SPSS applications (Xu et al., 2021).The https://penerbitadm.pubmedia.id/index.php/KOMITEKresults of problem identification at this stage are a description of the root of the problem regarding the population prediction system in South Bengkulu Regency and comparing the results of population prediction using python and SPSS applications (Alakus & Turkoglu, 2020).In addition, researchers made a description of the proposed solutions offered in this study (Wang et al., 2023).

Data Collection
Researchers collected data related to population prediction.The data collection methods used are interviews, surveys, and literature studies (Vásquez-Morales et al., 2019).

Data Processing
Researchers processed the data needed in further analysis, namely multiple linear regression tests on independent variables and related variables (Duffey & Zio, 2020).

Analysis
The researcher conducted a linear regression analysis of the independent and dependent variables related to the prediction of population in South Bengkulu Regency.
After conducting a linear regression test, variables that have a significant and insignificant effect on population prediction were obtained (Lebaz & Sheibat-Othman, 2019).

Inference
The researcher compiled the results and discussion of the comparison of the prediction of the population of South Bengkulu Regency using python and SPSS applications (Yan et al., 2020).

Documentation
This stage produces a research report that includes all activities that have been carried out by researchers (Ayati et al., 2021).

System Design
At this stage, researchers design web-based applications that implement the linear regression method in predicting population.The system design developed is flowchart and mock up.The system design is as follows.

Flowchart
The system design flow diagram is shown in Figure 3.1.

Implementation
In the implementation subchapter, there are several stages carried out, namely, data collection, data processing, regression modeling, testing results, and drawing conclusions.The stages that occur in this study are as follows (Afzal et al., 2023).

1) Data Collection
In the initial stage, it starts from preparing data that will be used as training data and test data in the formation of linear regression models.Population data to be analyzed is shown in Table 2.1.There are data columns for year, semester, population, birth, death, in-migration, and out-migration (Wannigamage et al., 2020).However, the data that will be used in multiple linear regression analysis is the total population (y) as the dependent variable, births (X1), deaths (X2), in-migration (X3), and out-migration (X4).Variable X is the independent variable (Hannon et al., 2023).In this research, there are manual calculations, SPSS calculations, and using python library calculations so that the data is made in (He et al., 2021).csv format which contains birth, death, in-migration, out-migration, and total population variables (Belsher et al., 2019).Furthermore, researchers can use this data to conduct multiple linear regression analysis using several methods of manual calculation, SPSS application assistance, and using the python library (Jaidka et al., 2020).

2) Analysis
Calculation Using Manual or Ms. Excel In the initial stages of manual calculation using Ms. Excel, researchers (Rabby et al., 2019) made a data filling table as an initial stage of creating a matrix in finding the regression equation shown in Table 2.2.The data that has been prepared is shown in Table 2.3 (Keuning et al., 2020).After filling in the data for the next matrix creation, researchers used the following formula to create the A matrix and H matrix shown in Table 2.4 and Table 2.5.The n is the amount of data analyzed.

Table 2
Matrix Preparation Data Filling

Table 3 .
Matrix A