Claim Missing Document
Check
Articles

KETEPATAN KLASIFIKASI KEIKUTSERTAAN KELUARGA BERENCANA MENGGUNAKAN REGRESI LOGISTIK BINER DAN REGRESI PROBIT BINER (Studi Kasus di Kabupaten Semarang Tahun 2014) Fajar Heru Setiawan; Rita Rahmawati; Suparti Suparti
Jurnal Gaussian Vol 4, No 4 (2015): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (364.408 KB) | DOI: 10.14710/j.gauss.v4i4.10219

Abstract

Population growth in Indonesia has increasedeach year. According to the population sensus conducted by National Statistics Bureau in 2010, Indonesia's population reached 237,6 million. Therefore, to control the population growth rate, government hold Keluarga Berencana (KB) or family planning program for couples in the childbearing age. The aim of this thesis which analyze the classification of couples in the childbearing age who follow family planning program, is to reduce the birth rate. So that, population can be controlled. The data used in this study is a Semarang Regency updated family data in 2014 that conducted Nasional Population and Family Panning Bureau. From the data, a binary logistic regression model and binary probit regression will be obtained, also classification accuracy will be obtained from each of these models. The analysis showed that the Binary Logistic Regression method produces a classification accuracy of 69,0% with 31,0% classification error. While, Probit Binary Regression method produces a classification accuracy of 68,4% with 31,6% misclassification. Binary Logistic Regression and Binary Logistic Regression method have a differences classification accuracy was very small then both are relative similar for analyze the classification family planning in Semarang Regency. Keywords: Keluarga Berencana (KB), Binary Logistic Regression, Binary Probit Regression, Classification,Confusion
REGRESI SPLINE SEBAGAI ALTERNATIF DALAM PEMODELAN KURS RUPIAH TERHADAP DOLAR AMERIKA SERIKAT Sulton Syafii Katijaya; Suparti Suparti; Sudarno Sudarno
Jurnal Gaussian Vol 2, No 3 (2013): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (652.651 KB) | DOI: 10.14710/j.gauss.v2i3.3668

Abstract

Exchange rate is the ratio of value or price of the currency between two countries. Many factors are thought to affect change in the inflation rate, the activity balance of payments, interest rate differentials, the relative level of income, government control and expectations. Therefore the method that can be used to analyze the exchange rate is needed such as the classical time series analysis (parametric). However the fluctuated data rate doesn’t occupy the assumption of stationarity often. Another alternative for this study is the spline regression. Spline is a nonparametric regression that doesn’t hold any assumption of regression curves. Spline regression has high flexibility and ability to estimate the data behavior which is likely to be different at every point of the interval, with the help of knots. The best model depends on the determination of the optimal point knots, that is has a minimum value of Generalized Cross Validation (GCV). Using data daily exchange rate of the rupiah against the dollar in the period of January 2, 2012 until October 15, 2012, the best spline model in this study is when using 2 to 3 order of approaching knots point, those points are 9512, 9517 and 9522 with the GCV = 1036.38.
PEMODELAN MULTIVARIATE ADAPTIVE REGRESSION SPLINES (MARS) PADA FAKTOR-FAKTOR RESIKO ANGKA KESAKITAN DIARE (Studi Kasus : Angka Kesakitan Diare Di Jawa Tengah, Jawa Timur Dan Daerah Istimewa Yogyakarta Tahun 2011) Wasis Wicaksono; Yuciana Wilandari; Suparti Suparti
Jurnal Gaussian Vol 3, No 2 (2014): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (621.283 KB) | DOI: 10.14710/j.gauss.v3i2.5913

Abstract

Diarrhea morbidity is a number of diarrhea suffers in specific region in period of one year per 1000 populations. Diarrhea morbidity is the impact from some factors such as environment, education, socioeconomic, nutrition and foods. Environmental factors that can affect the morbidity of diarrhea include the percentage of families who have a healthy latrine and percentage of households using clean water. For educational factors include the average length of school and literacy rates. On socio-economic factors include the percentage of poor and average people per household. While the food and nutritional factors are the percentage TUPM (Public Places and Food Management) healthy.Diarrhea morbidity can be pressed by analyzing those factors so that the prevention can be devised. Regression curve is used to draw the relationship of response variable and predictor variable and mostly approached by parametric regression, where the curve design is known (such as linear, quadratic and cubic). If curve design is unknown, then regression curve can be derived by approaching using non parametric regression. Multivariate Adaptive Regression Spline (MARS) is one of  nonparametric regression method that can be used on high dimension data. the best MARS model is derived by combination of Minimal Observation (MO), Maximum Basic Function (BF), and Maximal Interaction (MI) through trial and error. MARS model to predict diarrhea morbidity in Central Java, East Java and Yogyakarta is MARS (MO=2;BF=28;MI=3) and equation is  =  -0.526742 + 0.264444 * BF2 + 12.2382 * BF5 - 7.76719 * BF15 + 4.96445 * BF17
ANALISIS INDEKS HARGA SAHAM GABUNGAN (IHSG) DENGAN MENGGUNAKAN MODEL REGRESI KERNEL Icha Puspitasari; Suparti Suparti; Yuciana Wilandari
Jurnal Gaussian Vol 1, No 1 (2012): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (558.432 KB) | DOI: 10.14710/j.gauss.v1i1.577

Abstract

Saham merupakaninvestasi yang banyak dipilih para investor, salah satu indikator yang menunjukkan pergerakan harga saham adalah Indeks Harga Saham Gabungan (IHSG). IHSG merupakan data runtun waktu sehingga untuk menganalisisnya dapat menggunakan metode runtun waktu klasik. Namun dengan metode tersebut banyak asumsi yang harus dipenuhi, sehingga diperlukan metode alternatif salah satunya metode regresi nonparametrik karena dalam model regresi nonparametrik tidak ada asumsi khusus sehingga model ini merupakan metode alternatif yang dapat digunakan dalam analisis IHSG. Dalam makalah ini dibandingkan nilai MSE yang dihasilkan dari analisis runtun waktu klasik, regresi parametrik linier sederhana dan regresi nonparametrik kernel. Data IHSG yang digunakan adalah  periode minggu pertama Januari 2011 sampai dengan minggu ke empat Februari 2012. Data tersebut merupakan data closing price saham mingguan pada periode perdagangan terakhir. Hasil perbandingan nilai MSE dari dataIHSG yang sering fluktuatif pada tiga analisis didapatkan nilai MSE terkecil adalah pada analisis menggunakan regresi nonparametrik kernel dengan fungsi triangle dan badwidth h sebesar 58.2 dengan nilai MSE = 6987.787. Model terbaik tersebut dapat digunakan untuk memprediksikan nilai IHSG selanjutnya.
PERBANDINGAN METODE KLASIFIKASI NAÏVE BAYES DAN K-NEAREST NEIGHBOR PADA ANALISIS DATA STATUS KERJA DI KABUPATEN DEMAK TAHUN 2012 Riyan Eko Putri; Suparti Suparti; Rita Rahmawati
Jurnal Gaussian Vol 3, No 4 (2014): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (382.464 KB) | DOI: 10.14710/j.gauss.v3i4.8094

Abstract

Large population in Indonesia is closely related to the working status of the population which is unemployed or employed. It can lead to the high unemployment when the avaliable jobs arent balance with the population. Used two methods to perform the classification of employment status on the number of residents in the labor force in Demak for 2012 which is Naïve Bayes and K-Nearest Neighbor. Naïve Bayes is a classification method based on a simple probability calculation, while the K-Nearest Neighbor is a classification method based on the calculation of proximity. Variables used in determining whether a person's employment status is idle or not are gender, status in the household, marital status, education, and age. Employment status of the data processing methods of Naïve Bayes with the accuracy obtained is equal to 94.09% and the K-Nearest Neighbor method obtained is equal to 96.06% accuracy. To evaluate the results of the classification used calculations Press's Q and APER. Based on the analysis, the Press's Q values obtained indicate that both methods are already well in the classification of employment status data in Demak. Based on the calculation of APER, the classification of data in the employment status of Demak using the K-Nearest Neighbor method has an error rate smaller than the Naïve Bayes method. From this analysis it can be concluded that the K-Nearest Neighbor method works better compared with the Naïve Bayes for employment status data in the case of Demak for 2012. Keywords : Classification, Naïve Bayes, K-Nearest Neighbor (K-NN), Classification evaluation
PEMODELAN TINGKAT PENGANGGURAN TERBUKA DI JAWA TENGAH MENGGUNAKAN REGRESI SPLINE Seta Satria Utama; Suparti Suparti; Rita Rahmawati
Jurnal Gaussian Vol 4, No 1 (2015): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (715.47 KB) | DOI: 10.14710/j.gauss.v4i1.8151

Abstract

Unemployment is one of the employment problems facing Indonesia. Central Java Province is one of the provinces with a high enough unemployment. The main indicators used to measure the unemployment rate in the labor force that is unemployed. Based on research Arianie (2012) labor force participation rate significantly affect the unemployment rate and based on research Sari (2012) the gross enrollment ratio significantly affects the rate of open unemployment. Therefore, in this study using the two predictor variables with the labor force participation rate as X1 and gross enrollment rate as X2. This study aimed to explore the model of open unemployment rate in the Province of Central Java. The method used is the method of spline regression. Spline regression has the ability to adapt more effectively to the data patterns up or down dramatically with the help of dots knots. Determination of the optimal point knots are very influential in determining the best spline models. The best spline models are models that have a minimum GCV (Generalized Cross Validation) Value. Best spline models for the analysis of the data rate of unemployment in Central Java Province is the spline regression model when order X1 is 2 and order X2 is 4 and large number of knots in the X1 is 1 knot at the point 68.02394 and X2 is 3 knots at the point 82.13, 87.19, and 87.65 with GCV value of 1.732746. Keywords: Rate of  Open Unemployment, Spline Regression, GCV
ANALISIS KELOMPOK DENGAN ALGORITMA FUZZY C-MEANS DAN GUSTAFSON KESSEL CLUSTERING PADA INDEKS LQ45 Lailly Rahmatika; Suparti Suparti; Diah Safitri
Jurnal Gaussian Vol 4, No 3 (2015): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (432.087 KB) | DOI: 10.14710/j.gauss.v4i3.9478

Abstract

Clustering analysis is a data analysis aimed at determining a group of data based on common characteristics. Grouping method that’s being developed now is fuzzy clustering analysis. Fuzzy clustering algorithm that’s commonly used is the Fuzzy C-Means (FCM) algorithm and developed further by Gustafson Kessel Clustering (GK) which is able to detect groups with different shape than the FCM. This study examines the comparative application of FCM and GK clustering method in a case study, namely grouping in LQ45 based on the shares ratio of Earning Per Share (EPS) and Price Earning Ratio (PER). Determination of the optimal number of groups is done through calculation Xie and Beni validity index.In this research the algorithm FCM and GK will be made using MATLAB software, such as  GUI-based application program which can help users to perform clustering analysis. In some cases, the research results showed that GK is better than FCM, specifically in  generating the objective function and the standard deviation ratio of the minimum group. Based on the validity index Xie and Beni, it can be concluded that the optimal number of groups are divided into three.Keywords: Categories of Stocks, Fuzzy C-Means, Gustafson Kessel clustering, Xie and Beni index.
ANALISIS HUBUNGAN ANTARA LAMA STUDI, JALUR MASUK DAN INDEKS PRESTASI KUMULATIF (IPK) MENGGUNAKAN MODEL LOG LINIER (STUDI KASUS: LULUSAN MAHASISWA FSM UNDIP PERIODE WISUDA TAHUN 2012/2013) Diah Budiati; Yuciana Wilandari; Suparti Suparti
Jurnal Gaussian Vol 3, No 1 (2014): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (381.126 KB) | DOI: 10.14710/j.gauss.v3i1.4774

Abstract

Graduation is the end result of the process learning during the lectures in college. One of the duties and responsibilities of the college is to produce quality graduates, which college will prepare candidates reliable scholars, achievers and have special expertise in the field. To achieve S1 degree course each student must complete his college studies load. In the process of completion of the study load many factors at play, both internal and external factors. These factors are not directly specify a person in graduation. In this study, the internal factors are long study, driveways and university grade point average (GPA) of students. The purpose of this study was to determine the relationship between the internal factors in terms of graduation. One method used to determine the relationship between the factors is log linear models. Estimating a log linear model using the Maximum Likelihood Estimation (MLE), which is followed by Newton-Raphson iteration. Selection of the best model was conducted using Backward Elimination. To test the significance of the model has been obtained to use Goodness of Fit Test. After testing on the whole, it is known that each of the factors that play a role in graduate student tested and there was an interaction between the period of study with a GPA of factors.
Ketepatan Klasifikasi Status Pemberian Air Susu Ibu (ASI) Menggunakan Multivariate Adaptive Regression Splines (MARS) dan Algoritma C4.5 di Kabupaten Sragen Yusuf Arifka Rahman; Suparti Suparti; Sugito Sugito
Jurnal Gaussian Vol 5, No 1 (2016): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (502.401 KB) | DOI: 10.14710/j.gauss.v5i1.11062

Abstract

The progress of a nation influenced and determined by the level of public health, the indicator of the level of health is determined by nutritional status. Nutrition can be given early, namely breastfeeding to infants. This research aims to compare the classification of exclusive breastfeeding and nonexclusive breastfeeding. It used two methods for classifying a breastfeeding to babies in Sragen subdistrict on 2014, the methods are Multivariate Adaptive Regression Splines (MARS) and C4.5 Algorithm. MARS is nonparametric regression method that use to overcome the high dimension of data that produces accurate prediction and continuous models on knot. C4.5 Algorithm is a way of classifying methods from data mining that use to construct a decision tree. To evaluate the result of classification use Apparent Error Rate (APER) calculation. The best classification  result using MARS method is by using the combination of Basis Function (BF)=40, Maximum Interaction (MI)=3, Minimum Obsevation (MO)=3 because it will result on the smallest Generalized Cross Validation (GCV). Classification result using MARS method obtained APER is 19,7674% and 80,2326% of accuracy. Classification result using C4.5 Algorithm obtained APER is 18,6047% and 81,3953% of accuracy. From proportion test, concluded classification that formed by MARS is as good as by C4.5 Algorithm. Keywords: Breastfeeding, Classification, MARS, C4.5 Algorithm
KLASIFIKASI TINGKAT KELANCARAN NASABAH DALAM MEMBAYAR PREMI DENGAN MENGGUNAKAN METODE REGRESI LOGISTIK ORDINAL DAN NAÏVE BAYES (Studi Kasus pada Asuransi AJB Bumiputera Tanjung Karang Lampung) Ria Sutitis; Suparti Suparti; Dwi Ispriyanti
Jurnal Gaussian Vol 4, No 3 (2015): Jurnal Gaussian
Publisher : Department of Statistics, Faculty of Science and Mathematics, Universitas Diponegoro

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (341.885 KB) | DOI: 10.14710/j.gauss.v4i3.9489

Abstract

In the insurance companies a problem that often arises is the amount of customer debt in paying premiums, so it needs a system that can classify customers in the group not well, less smoothly, and smooth in paying premiums. Used two methods to perform the classification of payment premium status which is Regression Logistics Ordinal and Naïve Bayes. Variables used in determining whether a payment premium status are gender, marital status, age, work, income, insurance period, and the payment of premium. In Regression Logistics Ordinal, significant variables to the model are gender, marital status, age, insurance period, and the payment of premium. For significant variables used in the classification. Payment premium status of the data processing methods of Regression Logistics Ordinal with accuracy obtained is equal to 50.90% and the Naïve Bayes method obtained is equal to 55.41%. Based on the level of accuracy, the classification of data payment premium status of insurance AJB Bumiputera Tanjung Karang Lampung using the Naïve Bayes method has a greater degree of accuracy than the Regression Logistics Ordinal method. Keywords: Payment Premium Status, Classification, Naïve Bayes, Regression Logistics Ordinal