Abstract. This study aims to model the risk factors for maternal mortality in Southeast Sulawesi Province in 2022 using a negative binomial regression approach. This analysis is used to overcome the overdispersion problem in count data. The data used is secondary data obtained from the Southeast Sulawesi Provincial Health Service. The variables studied included the number of health centers, the percentage of births in health service facilities, the percentage of pregnant women who implemented the K4 program, the percentage of pregnant women who implemented the K1 program, and the number of midwives. The research results showed that the variables percentage of births in health service facilities, percentage of pregnant women implementing the K4 program, and number of midwives significantly affected maternal mortality. The negative binomial regression model illustrates that increasing the percentage of births in health service facilities and implementing the K4 program can reduce maternal mortality. On the other hand, increasing the number of midwives did not have a significant effect on reducing maternal mortality.

ABSTRAK Portofolio adalah kumpulan dari beberapa aset. Tujuan pembentukan portofolio yaitu untuk menghasilkan return yang paling tinggi sambil mengurangi risiko. Untuk menghindari outlier yang sering terjadi dalam portofolio model Mean-Variance (MV), perlu menggunakan estimasi robust. Data penelitian menggunakan closing price bulanan dari saham-saham yang konsisten selalu masuk ke dalam kelompok Jakarta Islamic Index (JII) periode Januari 2019 – Juni 2023. Tahap awal analisis menggunakan teknik klastering metode K-Medoids berdasarkan pada nilai expected return dan risiko. Klasterisasi dilakukan untuk menghemat waktu dan menekan biaya manajemen portofolio. Hasil analisis klaster menciptakan dua klaster. Saham INCO dan ADRO mewakili klaster pertama, dan saham KLBF dan AKRA mewakili klaster kedua. Keempat saham representasi tersebut dibentuk portofolio MV robust S dan portofolio MV robust Constrained-M (CM). Kinerja portofolio diukur menggunakan sharpe ratio. Hasil analisis menunjukkan bahwa kinerja model portofolio robust MV estimasi Constrained-M (CM) mengungguli kinerja model portofolio robust MV estimasi S. ABSTRACT A portfolio is an assortment of several items. The goal of portfolio construction is to get the maximum return at the least amount of risk. Robust estimate is a means to mitigate the sensitivity of the Mean-Variance (MV) model portfolio to outliers. The research data uses monthly closing prices of stocks that are consistently included in the Jakarta Islamic Index (JII) group for the period January 2019 - June 2023. The initial stage of analysis uses the K-Medoids method clustering technique based on the expected return and risk values. The purpose of the clustering is for time efficiency and to reduce the amount of costs in managing the portfolio. The results of the cluster analysis formed two clusters, where the first cluster is represented by INCO and ADRO stocks. While the second cluster is represented by KLBF dan AKRA stocks. The four representative stocks are formed MV robust S portfolio and MV robust Constrained-M (CM) portfolio. Portfolio performance is measured using the sharpe ratio. According to the analysis's findings, the Constrained-M (CM) estimation MV robust portfolio model performs better than the S estimation MV robust portfolio model.

ABSTRAK Data pertumbuhan GDP empat negara asia timur tidak hanya bergantung pada waktu (time), tetapi juga bergantung pada tempat (space). Model space time merupakan metode yang populer digunakan dalam peramalan untuk menganalisis data time series yang mempertimbangkan faktor time dan place secara bersamaan. Dalam hal pengambilan parameter, model untuk GSTARIMA dan model untuk STARIMA berbeda. Untuk model STARIMA, menggunakan lokasi penelitian yang sama, sementara model GSTARIMA digunakan untuk lokasi penelitian yang tidak sama (heterogen) dan ditunjukkan dalam matriks pembobot. Tujuan penelitian ini adalah untuk menemukan model terbaik antara GSTARIMA atau STARIMA guna memodelkan pertumbuhan GDP dari beberapa negara di Asia timur yakni Indonesia, China, Japan, dan Korea Selatan. Interval data dimulai dari tahun 1962 hingga tahun 2017. Karena terdapat hubungan yang signifikan, pemodelan menggunakan STARIMA dan GSTARIMA diperlukan untuk mengevaluasi pertumbuhan GDP. Kedua model ini dikembangkan melalui pendekatan VAR dengan orde spasial satu, yang mencakup konfigurasi STAR(1,1) dan GSTAR(1,1) masing-masing. Hasil pengujian menunjukkan bahwa model terbaik untuk meramalkan peningkatan GDP di masa depan adalah STAR(1), dengan nilai RMSE terkecil. Melalui pemodelan ini dapat meramalkan kondisi GDP di masa mendatang dan mengevaluasi apakah hubungan internasional terutama dalam hal kerjasama ekonomi yang telah dijalankan pada ke empat negara berdampak positif terhadap pertumbuhan ekonomi. ABSTRACT The GDP growth data of four East Asian countries depends not only on time but also on space. The space-time model is a popular method used in forecasting to analyze time series data while simultaneously considering both time and spatial factors. In terms of parameter selection, the models for GSTARIMA and STARIMA differ. The STARIMA model uses the same research location, while the GSTARIMA model is applied to heterogeneous research locations, as represented in the weighting matrix. The purpose of this study is to determine the best model between GSTARIMA and STARIMA to model GDP growth in several East Asian countries, namely Indonesia, China, Japan, and South Korea. ...

ABSTRAK After Sales memiliki kontribusi positif pada pertumbuhan profit permintaan spare part suatu perusahaan. PT. XYZ merupakan agen tunggal pemegang merek mobil kendaraan niaga di Indonesia dengan visi menjadi market share terbesar secara global dipasar internasional. Peranan penting yang dilakukan PT. XYZ dalam upaya memenuhi seluruh kebutuhan permintaan domestik atau ekspor dengan langkah menerapkan rencana inventaris yang mencakup peramalan kebutuhan suku cadang setiap bulan. Pada kondisi saat ini metode yang digunakan oleh perusahaan hanya menggunakan metode peramalan dengan rata-rata 6 bulan terakhir atau Moving Average yang digunakan untuk menyiapkan seluruh suku cadang PT. XYZ. Metode ini dinilai kurang efektif karena ketika terjadinya bentuk permintaan spare part yang memiliki tingkat variasi tinggi maka penyimpangan pada hasil peramalan menyebabkan back order terhadap pelanggan dan loss sales yang dapat mempengaruhi nilai suatu perusahaan yaitu fast and easy process. Berdasarkan permasalahan tersebut, maka PT. XYZ membutuhkan perbaikan terhadap mekanisme peramalan spare part-nya. Tujuan dari dilakukan penelitian yaitu menghasilkan metode klasifikasi spare part berdasarkan bentuk permintaannya dan memutuskan metode peramalan yang paling tepat untuk masing-masing kelompok spare part dengan langkah membandingkan enam metode peramalan yaitu metode Croston Optimized, TSB, SESOpt, ADIDA, IMAPA, Moving Average. Semua metode peramalan akan dibandingkan atas dasar parameter nilai forecasting error. Pada hasil yang didapat menunjukkan metode Croston Optimized memiliki hasil yang lebih baik dari metode moving average dengan memperbaiki kesalahan sebesar 7%. ABSTRACT After Sales has a positive contribution to the profit growth of a company's spare part demand. PT XYZ is the sole agent of commercial vehicle car brand holders in Indonesia with a vision to become the largest market share globally in the international market. An important role played by PT XYZ in an effort to meet all domestic or export demand needs by implementing an inventory plan that includes forecasting the need for spare parts on a monthly basis. In the current condition, the method used by the company only uses a forecasting method with the average of the last 6 months or Moving Average which is used to prepare all spare parts of PT. XYZ. This method is considered less effective because when there is a form of spare part demand that has a high level of variation, deviations in forecasting results cause back orders to customers and loss of sales which can affect the value of a company, namely fast and easy process. Based on these problems, PT XYZ needs improvements to its spare part forecasting mechanism. The purpose of the research is to produce a spare part classification method based on the form of demand and decide the most appropriate forecasting method for each spare part group by comparing six forecasting methods, namely the Croston Optimized method, TSB, SESOpt, ADIDA, IMAPA, Moving Average. All forecasting methods will be compared on the basis of the forecasting error value parameter. The results obtained show that the Croston Optimized method has better results than the moving average method by correcting an error of 7%.

ABSTRAK Stres adalah respons adaptif individual terhadap stimulus yang dianggap mengancam, dapat dialami oleh siapa saja, termasuk mahasiswa tingkat akhir. Penelitian ini fokus pada faktor-faktor yang memengaruhi tingkat stres mahasiswa tingkat akhir S1 Matematika di Universitas Negeri Medan, dengan penekanan pada penyusunan tugas akhir (skripsi). Beberapa faktor, seperti tuntutan akademik, masalah pribadi (manajemen waktu, motivasi, dukungan keluarga), dan aturan batas masa studi, diidentifikasi sebagai penyebab stres.Metode penelitian menggunakan pendekatan kuantitatif non-eksperimental dengan uji korelasi Rank Spearman. Hasil penelitian menunjukkan korelasi cukup kuat (0.380) antara faktor kejadian stres dan tingkat stres mahasiswa. Meskipun tidak mencapai tingkat signifikansi yang umum, temuan ini mengindikasikan bahwa mahasiswa yang mengalami berbagai faktor kejadian stres cenderung memiliki tingkat stres yang lebih tinggi. Penelitian ini memberikan wawasan mendalam tentang kompleksitas faktor-faktor yang berkontribusi terhadap stres mahasiswa tingkat akhir, menunjukkan perlunya pendekatan holistik dan intervensi terintegrasi dari perguruan tinggi untuk membantu mahasiswa mengelola stres dengan lebih efektif. Temuan ini juga memberikan dasar bagi pengembangan program dukungan dan pembinaan untuk membantu mahasiswa mengatasi tantangan akademik dan pribadi. ABSTRACT Stress is an individual adaptive response to a stimulus that is perceived as threatening, can be experienced by anyone, including final year students. This research focuses on the factors that influence the stress level of final year undergraduate mathematics students at Medan State University, with an emphasis on the preparation of the final project (thesis). Several factors, such as academic demands, personal problems (time management, motivation, family support), and study period limit rules, were identified as causes of stress.The research method used a non-experimental quantitative approach with the Spearman Rank correlation test. The results showed a moderately strong correlation (0.380) between stress event factors and students' stress levels. Although not reaching the usual level of significance, this finding indicates that college students who experience multiple stressful event factors tend to have higher stress levels. This study provides an in-depth insight into the complexity of factors contributing to final year university students' stress, suggesting the need for a holistic approach and integrated interventions from universities to help students manage stress more effectively. The findings also provide a basis for the development of support and coaching programs to help college students overcome academic and personal challenges.

ABSTRACT Fresh and chilled fish are the largest contributor to Indonesia's fishery product exports, accounting for a share of 45% or around USD 2.2 billion in 2021. The main destination countries for Indonesia's fresh and chilled fish exports include China, the United States, Japan, and other European countries. This research aims to analyze the factors influencing the export value of Indonesia's fresh and chilled fish, as well as to identify and evaluate the ARIMA (p, d, q) model based on historical data from 2012-2022. The result is an ARIMA (4,2,2) model with a MAPE value of 2.208 and a predicted value for the 2023 period of 4351 tons. This is in line with the large exports of fresh fish from Indonesia to various destination countries.

ABSTRACT Seismic hazard expressed in terms of annual rate of exceedance and is used to calculate the earthquake insurance premium. Annual rate of exceedance is a complicated function of magnitudes, distances from site to earthquake sources and attenuation. Due to its complexity, determination of exact sampling distribution of earthquake insurance premium is not an easy task. This research proposes Monte Carlo simulation approach to determine the sampling distribution of earthquake insurance premium. Annual rate of exceedance was simulated first and then the insurance premium calculated based on simulation of annual rate of exceedance. The simulation involves quantifying synthetic catalogue similar to historical catalogue. Its simulation is conducted in order to construct annual rate of exceedance as an indicator of earthquake risk used in earthquake insurance. The simulation of 25 iteration and sample size of 100 shows that the sampling distribution of insurance premium is skewed to the right. The idea of using Monte Carlo simulation to study the sampling distribution of annual rate of exceedance is the originality of the study and the study contributes to the methodology of earthquake insurance.

ABSTRAK Daya simpan panas (DSP) gerabah vorno merupakan suatu kebutuhan untuk meningkatkan produksi sagu lempeng. Metode eksprimen penambahan aluvial pada pembuatan gerabah dengan komposisi tanah liat 40 %, aluvial 25% dan pasir pantai 35% sebagai gerabah pengembangan diharapkan dapat meningkatkan DSP gerabah. Keterbatasan instrumen pengujian DSP secara teknik, membutuhkan solusi secara statistik dalam mendeterminasi DSP. Perbandingan hasil analisis dari pendekatan metode linear dan non linear pada data perubahan suhu lempengan membentuk karakteristik tertentu, olehnya itu dibutuhkan identifikasi analisis statistik yang tepat dalam mendeterminasi DSP dengan akurasi tinggi. Hasil penelitian menunjukkan bahwa penggunaan material aluvial pada pembuatan gerabah meningkatkan berat jenis lempengan gerabah. Koefisien determinasi (R2) dari keseluruhan persamaan metode linear dan non linear memberikan linearity yang tinggi (R2 > 75%), akurasi tertinggi dengan nilai mean absolute percent error (MAPE) < 10% dijumpai pada tren persamaan kubik dan logaritmik. Pendekatan tren kubik menunjukkan DSP gerabah pengembangan lebih lambat 7% dibandingkan dengan gerabah rujukan, sedangkan penggunaan tren logaritik lebih lambat 33%. Perlu dilakukan analisis data dengan perhitungan transfer panas untuk membuktikan DSP dengan pendekatan statistik. ABSTRACT The heat storage capacity (HSC) of vorno pottery is a necessity to increase the production of sago plates. The experimental method of adding alluvial to pottery making with a composition of 40% clay, 25% alluvial and 35% beach sand as development pottery is expected to increase the DSP of pottery. The limitations of technical DSP testing instruments require a statistical solution in determining DSP. Comparison of analysis results from linear and non-linear method approaches on slab temperature change data forms a specific characteristic, therefore it is necessary to identify the right statistical analysis in determining DSP with high accuracy. . The results show that the use of alluvial materials in pottery making increases the specific gravity of pottery slabs. The coefficient of determination (R2) of all equations of linear and non-linear methods gave high linearity (R2 > 75%), the highest accuracy with mean absolute percent error (MAPE) value < 10% was found in the cubic and logarithmic trend equations. The cubic trend approach showed that the heat retention of the development pottery was 7% slower than the reference pottery, while the use of the logarithmic trend was 33% slower. It is necessary to analyse the data with heat transfer calculations to prove the HSC with a statistical approach.

ABSTRAK Dalam era pertumbuhan ekonomi dan kemajuan teknologi yang pesat, kebutuhan akan pasokan listrik yang konsisten semakin mendesak. Penelitian ini membandingkan metode prediksi error trend seasonality (ETS) dan prediksi linear untuk mengantisipasi ketersediaan listrik di Jawa Tengah, menggunakan data dari 2011 hingga 2021. Evaluasi menunjukkan bahwa metode prediksi linear lebih akurat dengan persentase kesalahan lebih rendah dibandingkan error trend seasonality (ETS). Prediksi menunjukkan tren positif peningkatan ketersediaan listrik, membuka peluang pengembangan industri dan pertumbuhan ekonomi regional. Keandalan prediksi linear memberikan dasar yang solid untuk pengambilan keputusan di sektor energi. Rekomendasi untuk meningkatkan keandalan prediksi di masa depan mencakup pemantauan dan pembaruan data, serta melibatkan pemangku kepentingan dan ahli energi dalam pengambilan keputusan. Penelitian ini diharapkan memberikan panduan bagi pembuat kebijakan dan pihak terkait dalam mengelola ketersediaan energi listrik secara efisien di Jawa Tengah. ABSTRACT In an era of rapid economic growth and technological progress, the need for a consistent electricity supply is increasingly urgent. This research compares the error trend seasonality prediction (ETS) method and linear prediction to anticipate electricity availability in Central Java, using data from 2011 to 2021. The evaluation shows that the linear prediction method is more accurate with a lower error percentage than error trend seasonality (ETS). Predictions show a positive trend of increasing electricity availability, opening up opportunities for industrial development and regional economic growth. The reliability of linear predictions provides a solid basis for decision making in the energy sector. Recommendations to improve the reliability of future predictions include monitoring and updating data, as well as involving stakeholders and energy experts in decision making. This research is expected to provide guidance for policy holders and related parties in managing the availability of electrical energy efficiently in Central Java.