Claim Missing Document
Check
Articles

Found 8 Documents
Search

Analisis Lintas Sifat Morfo-Agronomis dan Fisiologis Jagung (Zea mays L.) Annisa Malik; Farit Mochamad Afendi; Akbar Rizki; . Sutoro
Xplore: Journal of Statistics Vol. 2 No. 1 (2018): 30 Juni 2018
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (937.261 KB) | DOI: 10.29244/xplore.v2i1.72

Abstract

Corn (Zea mays L.) is the third most important food commodity after wheat and rice based on the world's staple food, and ranks second after rice based on staple food in Indonesia. High yielding varieties of corn are highly needed to meet food, feed and industrial needs. These varieties can be obtained through plant breeding programs by utilizing the source of genes capable of producing good plant character. Gene sources can be obtained from germplasm or local varieties that exist. Character of plants that can support the productivity of plants can be used as an indicator of the selection process in corn plant breeding. This can be done through characterization of morpho-agronomic and physiological properties of each corn variety, then determine the characters that support the productivity of corn plants directly or indirectly. The direct and indirect effect of a plant's character on crop productivity is identified through path analysis. The results showed that the effective selection criteria for increasing corn’s seed weight directly was the leaf area. While the effective selection criteria for increasing the weight of biomass directly is the age of female flowers out. While the effective selection criteria for increasing corn’s seed weight and biomass indirectly is the plant height through the filling rate of the seeds.
Penggerombolan Provinsi di Indonesia Berdasarkan Produktivitas Tanaman Pangan Tahun 2005-2015 Menggunakan Metode K-Error Emeylia Safitri; I Made Sumertajaya; Akbar Rizki
Xplore: Journal of Statistics Vol. 2 No. 1 (2018): 30 Juni 2018
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (450.475 KB) | DOI: 10.29244/xplore.v2i1.75

Abstract

Clustering analysis is a multivariate analysis that’s aim for gruping the observasion objects to some groups. The clusters have low similarity between the clusters and high similarity in same cluster. Classic grouping analysis have a weakness that doesn’t insert measurement error information that related with data. Clustering analysis with K-Error method is expanded for solusing solving the measurement error data problem in classic grouping analysis. The research is aim for clustering the provinces in Indonesia using K-Error and K-Means method based on crops productivity. K-Error method produces better clusters than KMeans. K-Error method formed 7 clusters. Cluster 5 consist of provinces with highest productivity almost at all crops. Cluster 2 and 3 have low productivity for partial crops.
Penanganan Data Tidak Seimbang pada Pemodelan Rotation Forest Keberhasilan Studi Mahasiswa Program Magister IPB Junjun Wijaya; Agus M Soleh; Akbar Rizki
Xplore: Journal of Statistics Vol. 2 No. 2 (2018): 31 Agustus 2018
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (234.149 KB) | DOI: 10.29244/xplore.v2i2.99

Abstract

Graduate school of Bogor Agricultural University (SPs-IPB) stated that not all students of IPB master program successfully complete their studies. This becomes an evaluation for IPB to be more selective in choosing students in the future. This study aims to model the success classification of IPB master students in 2011 to 2015. The classification method used is rotation forest. The percentage of students who graduated is very large compared to those who did not pass, this can cause the evaluation value different. SMOTE (Synthetic Minority Oversampling Technique) is one of method to handle such unbalanced data by generating artificial data. The ROC (Receiver Operating Characteristic) curve is built to see the optimum cut off value. There are two classification models, they are rotation forest models before and after handled by SMOTE. The comparison results show that the rotation forest model after SMOTE with cut off value 0.6 is the best model. This model can increase the sensitivity value more than 50% although the accuracy and specificity value decreased compared to the modeling before SMOTE.
PEMODELAN SEMIPARAMETRIK STATISTICAL DOWNSCALING UNTUK MENDUGA CURAH HUJAN BULANAN DI INDRAMAYU Akbar Rizki; Abdul Aziz Nurussadad
Xplore: Journal of Statistics Vol. 2 No. 2 (2018): 31 Agustus 2018
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (287.299 KB) | DOI: 10.29244/xplore.v2i2.117

Abstract

Semiparametric statistical downscaling (SD) model is a statistical model which consists of parametric and non-parametric functional relationship between local scale and global scale variable. This study used rainfall intensity in Indramayu as local scale variable and Global Precipitation Climatology Project (GPCP) precipitation as global scale variable. GPCP precipitation data have multicollinearity, therefore they were reduced by principal component analysis. Eight principal components which have been selected then used as the prediktors and rainfall intensity in Indramayu as the response. Semiparametric SD model was used to predict the rainfall intensity in the district of Indramayu. The semiparametric model developed by mixed model approach where the nonparametric relationship is represented using spline with truncated power basis. Linier semiparametric model is the best model to estimate monthly rainfall in indramayu district. The model performance evaluated by RMSEP (root mean square error prediction) and (coefficient of determination). The result shows that the best model have values of RMSEP and are 61.64 and 71%.
Penanganan Overdispersi pada Regresi Poisson dengan Regresi Binomial Negatif pada Kasus Kemiskinan di Indonesia Lulu Mahdiyah Sandjadirja; Muhammad Nur Aidi; Akbar Rizki
Xplore: Journal of Statistics Vol. 8 No. 1 (2019): 30 April 2019
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29244/xplore.v8i1.165

Abstract

Poisson regression can be used to model rare events that consist of count data. Poisson regression application is carried out to find out external factors that affect the number of poor people in Indonesia by the province in 2016. The assumptions that must be met in this analysis are equdispersion. However, in real cases there is often a problem of overdispersion, ie the value of the variance is greater than the average value. High diversity can be caused by outliers. Expenditures on outliers have not been able to deal with the problem of overdispersion in Poisson Regression. One way to overcome this problem is to replace the Poisson distribution assumption with the Negative Binomial distribution. The results of the analysis show that the Negative Binomial Regression model without outliers is better than the Poisson Regression without outliers model indicated by a smaller AIC value. Based on the Negative Binomial Regression model without this outlier the external factors that affect the number of poor people in Indonesia by the province in 2016 are the percentage of households with floor conditions of houses with soil by province, population by province, percentage of unemployment to the total workforce by province and the percentage of the workforce against the working age population.
Penerapan Regresi Logistik Biner Multilevel terhadap Ketepatan Waktu Lulus Mahasiswa Program Magister Sekolah Pascasarjana IPB Zana Aprillia; Farit Mochamad Afendi; Akbar Rizki
Xplore: Journal of Statistics Vol. 10 No. 2 (2021)
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (288.036 KB) | DOI: 10.29244/xplore.v10i2.199

Abstract

The study length of alumnus is one of the study achievement indicator of the university. Study length for Master Program can be divided into two categories which is pass on time (study length ≤24 months) and pass not on time (study length >24 months). In the classical regression analysis, each student are assumed to be independent. But in reality, each student are grouped into a different study programs so that the individuals who are in the same study program tend to have a similar characteristics. Multilevel regression is one of the analysis that accomodates the problem. The level used in this study are level 1 (individual student) and level 2 (study programs). The best multilevel regression model obtained is a model with random intercept and the variance is produced from study program is 0.6636. Factors that give an effect to the graduation’s timeliness are age, married status, and the source of the S2 education cost.
Penerapan Synthetic Minority Oversampling Technique pada Pemodelan Regresi Logistik Biner terhadap Keberhasilan Studi Mahasiswa Program Magister IPB Mega Pradita Pangestika; I Made Sumertajaya; Akbar Rizki
Xplore: Journal of Statistics Vol. 10 No. 2 (2021)
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (805.99 KB) | DOI: 10.29244/xplore.v10i2.238

Abstract

The Postgraduate School of IPB has academic standards as well as high competitiveness of graduates who have spread both at home and abroad. In this study Binary Logistic Regression method was used to determine the factors that influence the success of the study of Postgraduate students of Bogor Agricultural University (Graduate School-IPB). The data used are data from IPB Graduate School students who graduated from 2011 to 2015. The response variable used is the success status of student studies namely graduating and not passing and using 9 explanatory variables namely gender, marital status, admission status when entering S2, college status S1 level, the source of S2 education costs, group of agencies working, S2 study program groups, age when entering S2 and S1 GPA. The data obtained is not balanced with the percentage of students who graduate is greater than those who did not pass, so the imbalance of data is handled with SMOTE if it is not handled it will cause a misclassification. Comparison of classification results seen in testing data. The results in the model before SMOTE have an area under the curve or AUC of 0.6760, an accuracy value of 88.77%, a sensitivity value of 99.09% and a specificity of 4.63%. The model after 600% oversampling SMOTE has an AUC value of 0.6642, an accuracy value of 78.36%, a sensitivity value of 83.65%, and a specificity value of 35.18%. Although the accuracy of the model and sensitivity value before SMOTE was higher than the model after SMOTE, the specificity in the model after SMOTE was higher, which meant that the model after SMOTE was better at predicting minority classes (not graduating).
Analisis Kepuasan Terhadap Green Transportation Salvina Salvina; Akbar Rizki; Indahwati Indahwati
Xplore: Journal of Statistics Vol. 9 No. 1 (2020)
Publisher : Department of Statistics, IPB

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (243.654 KB) | DOI: 10.29244/xplore.v9i1.251

Abstract

ABSTRACT SALVINA. Analysis of Satisfaction against Green Transportation. Supervised by AKBAR RIZKI and INDAHWATI. One of the stages of the Green Campus 2020 program as an effort of IPB towards World Class University (WCU) is to carry out the Green Transportation (GT) movement. Buses, electric cars, bicycles and electric motorcycle taxis are the GT transportation modes in IPB. The purpose of this study was to determine the level of satisfaction of GT users and identify attributes that are important and need to be improved so that the GT service system can be improved. This study uses survey data conducted by researchers on undergraduate students who use GT transportation mode in the past week. The sampling method used is random layered sampling with layers in the form of faculties. The analytical methods used are Importance Performance Analysis (IPA), Customer Satisfaction Index (CSI), biplot analysis, and simple correspondence analysis. The CSI value obtained is 2.96 (1-4 scale) with a CSI percentage of 74% in other words the user is satisfied with the service he receives. The aspects that need to be improved are aspects of empathy and reliability on electric cars and assurance on bicycles, while other aspects have been considered good. Biplot analysis shows the diversity of satisfaction obtained from aspects (reliability, empathy, tangibles, assurance, and responsiveness) is the same. Simple correspondence analysis shows students of the Faculty of Veterinary Medicine (FKH), Faculty of Animal Husbandry (FAPET), Faculty of Forestry (FAHUTAN), and General Competency Education Program (PPKU) often use electric cars. Faculties that often use buses are Faculty of Agriculture (FAPERTA), Faculty of Agricultural Technology (FATETA), Faculty of Fisheries and Marine Sciences (FPIK) and Faculty of Mathematics and Natural Sciences (FMIPA). The mode of bicycle transportation cannot be characterized in any faculty because at least the respondents use it. Keywords: biplot, green transportation, IPA-CSI, simple correspondence