Statistical Learning for Predicting Dengue Fever Rate in Surabaya
Dengue fever happening most in tropical countries and considered as the fastest spreading mosquito-borne disease which is endemic and estimated to have 96 million cases annually. It is transmitted by Aedes mosquito which infected with a dengue virus. Therefore, predicting the dengue fever rate as be...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Petra Christian University
2020-06-01
|
| Series: | Jurnal Teknik Industri |
| Subjects: |
| _version_ | 1798045204630994944 |
|---|---|
| author | Siana Halim Felecia Tanti Octavia |
| author_facet | Siana Halim Felecia Tanti Octavia |
| author_sort | Siana Halim |
| collection | DOAJ |
| description | Dengue fever happening most in tropical countries and considered as the fastest spreading mosquito-borne disease which is endemic and estimated to have 96 million cases annually. It is transmitted by Aedes mosquito which infected with a dengue virus. Therefore, predicting the dengue fever rate as become the subject of researches in many tropical countries. Some of them use statistical and machine learning approach to predict the rate of the disease so that the government can prevent that incident. In this study, we explore many models in the statistical learning approaches for predicting the dengue fever rate. We applied several methods in the predictive statistics such as regression, spatial regression, geographically weighted regression and robust geographically weighted regression to predict the dengue fever rate in Surabaya. We then analyse the results, compare them based on the mean square error. Those four models are chosen, to show the global estimator’s approaches, e.g. regression, and the local ones, e.g. geographically weighted regression. The model with the minimum mean square error is regarded as the most suitable model in the statistical learning area for solving the problem. Here, we look at the estimates of the dengue fever rate in the year 2012, to 2017, area, poverty percentage, precipitation, number of rainy days for predicting the dengue fever outbreak in the year 2018. In this study, the pattern of the predicted model can follow the pattern of the true dataset. |
| first_indexed | 2024-04-11T23:16:56Z |
| format | Article |
| id | doaj.art-b24a3ef2e4684a79b25cca6a0a58e2f1 |
| institution | Directory Open Access Journal |
| issn | 1411-2485 |
| language | English |
| last_indexed | 2024-04-11T23:16:56Z |
| publishDate | 2020-06-01 |
| publisher | Petra Christian University |
| record_format | Article |
| series | Jurnal Teknik Industri |
| spelling | doaj.art-b24a3ef2e4684a79b25cca6a0a58e2f12022-12-22T03:57:36ZengPetra Christian UniversityJurnal Teknik Industri1411-24852020-06-012213746https://doi.org/10.9744/jti.22.1.37-46Statistical Learning for Predicting Dengue Fever Rate in SurabayaSiana Halim0Felecia1Tanti Octavia2Industrial Engineering Department Petra Christian UniversityIndustrial Engineering Department Petra Christian UniversityIndustrial Engineering Department Petra Christian UniversityDengue fever happening most in tropical countries and considered as the fastest spreading mosquito-borne disease which is endemic and estimated to have 96 million cases annually. It is transmitted by Aedes mosquito which infected with a dengue virus. Therefore, predicting the dengue fever rate as become the subject of researches in many tropical countries. Some of them use statistical and machine learning approach to predict the rate of the disease so that the government can prevent that incident. In this study, we explore many models in the statistical learning approaches for predicting the dengue fever rate. We applied several methods in the predictive statistics such as regression, spatial regression, geographically weighted regression and robust geographically weighted regression to predict the dengue fever rate in Surabaya. We then analyse the results, compare them based on the mean square error. Those four models are chosen, to show the global estimator’s approaches, e.g. regression, and the local ones, e.g. geographically weighted regression. The model with the minimum mean square error is regarded as the most suitable model in the statistical learning area for solving the problem. Here, we look at the estimates of the dengue fever rate in the year 2012, to 2017, area, poverty percentage, precipitation, number of rainy days for predicting the dengue fever outbreak in the year 2018. In this study, the pattern of the predicted model can follow the pattern of the true dataset.global moran i statisticslocal moran i statisticsregressionspatial regressiongeographically weighted regression |
| spellingShingle | Siana Halim Felecia Tanti Octavia Statistical Learning for Predicting Dengue Fever Rate in Surabaya Jurnal Teknik Industri global moran i statistics local moran i statistics regression spatial regression geographically weighted regression |
| title | Statistical Learning for Predicting Dengue Fever Rate in Surabaya |
| title_full | Statistical Learning for Predicting Dengue Fever Rate in Surabaya |
| title_fullStr | Statistical Learning for Predicting Dengue Fever Rate in Surabaya |
| title_full_unstemmed | Statistical Learning for Predicting Dengue Fever Rate in Surabaya |
| title_short | Statistical Learning for Predicting Dengue Fever Rate in Surabaya |
| title_sort | statistical learning for predicting dengue fever rate in surabaya |
| topic | global moran i statistics local moran i statistics regression spatial regression geographically weighted regression |
| work_keys_str_mv | AT sianahalim statisticallearningforpredictingdenguefeverrateinsurabaya AT felecia statisticallearningforpredictingdenguefeverrateinsurabaya AT tantioctavia statisticallearningforpredictingdenguefeverrateinsurabaya |