Identifikasi Gagal Ginjal Kronis dengan Mengimplementasikan Metode Support Vector Machine beserta K-Nearest Neighbour (SVM-KNN)

Penulis

  • Alvin Tarisa Akbar Universitas Brawijaya, Malang
  • Novanto Yudistira Universitas Brawijaya, Malang
  • Achmad Ridok Universitas Brawijaya, Malang

DOI:

https://doi.org/10.25126/jtiik.20231026059

Abstrak

Ginjal merupakan bagian vital bagi manusia karena berfungsi untuk menyaring atau membersihkan cairan yang kita minum agar dapat dikonsumsi oleh tumbuh secara normal. Gagal ginjal adalah situasi dimana ginjal mengalami penurunan funsionalnya secara terus-menerus yang mana dapat mengakibatkan ketidakmampuan ginjal untuk berfungsi untuk semestinya. Untuk membantu pasien yang terjangkit penyakit gagal ginjal kronis hal yang terlebih dahulu dilakukan adalah mengindentifikasi penyakit tersebut. Indentifikasi gagal ginjal kronis dengan menggunakan dataset yang dibuat oleh L.Jerlin Rubini dkk. sudah dilakukan dengan berbagai metode klasifikasi, contohnya adalah implementasi metode klasifikasi Support Vector Machine (SVM) dan K-Nearest Neighbour (KNN). Salah satu kelemahan dari SVM adalah bila data terlalu dekat dengan hyperplane adanya potensi untuk salah mengklasifikasi. Lalu salah satu kelemahan dari KNN adalah berpotensi mengalami penuruan akurasi bila nilai k terlalu tinggi atau terlalu rendah yang mana masing-masing mengakibatkan banyaknya noise data atau terlalu kecil data yang digunakan sebagai pembanding. Untuk penelitian ini, kami mengimplementasikan penggabungan metode SVM dengan KNN yang dikenal dengan SVM-KNN yang menggunakan optimasi Simplified Sequential Minimal Optimization (Simplified SMO). Metode ini mencoba untuk menutupi kelemahan dari SVM dan KNN. Penelitian ini melakukan percobaan pada beberapa nilai parameter yang digunakan untuk mendapatkan akurasi pada metode klasifikasi SVM-KNN terbaik.  Parameter yang diuji adalah cost, tolerance, gamma, dan bias pada metode SVM, parameter k pada metode KNN, serta parameter miu pada metode SVM-KNN. Nilai rata-rata akurasi terbaik didapatkan dengan menggunakan SVM-KNN dengan nilai 94,25% dan terbukti lebih baik dari pada SVM dengan 94,09% dan KNN dengan 91,73%.

 

Abstract


Kidneys are a vital part for humans because they function to filter or clean the fluids we ingest so that they can be consumed safely. Kidney failure is a situation where the kidneys experience a continuous decline in function which can result in the inability of the kidneys to function properly. To help patients with chronic kidney failure, the first thing to do is to identify the disease. Identification of chronic kidney failure using the dataset created by L.Jerlin Rubini et. al. had been tested with various classification methods, for example the implementation of the Support Vector Machine (SVM) and K-Nearest Neighbor (KNN). One of the weaknesses of SVM is that if the data is too close to the hyperplane there is the potential for misclassification. Then one of the weaknesses of KNN is that it has the potential to experience a decrease in accuracy if the value of k is too high or too low which results in a lot of noise data or too little data used as a comparison respectively. For this research, we implemented a hybrid of SVM with KNN known as SVM-KNN which was optimized using Simplified Sequential Minimal Optimization (Simplified SMO). This study conducted experiments on several parameter values used to obtain the best accuracy in SVM-KNN. The parameters tested are cost, tolerance, gamma, bias on SVM, parameter k on KNN, and miu on SVM-KNN. The average value of accuracy was obtained using SVM-KNN with 94.25% and proved better than SVM with 94.09% and KNN with 91.73%.

Downloads

Download data is not yet available.

Referensi

AHMAD, M., TUNDJUNGSARI, V., WIDIANTI, D., AMALIA, P., & RACHMAWATI, U. 2017. Diagnostic Decision Support System of Chronic. Second International Conference on Informatics and Computing (ICIC).

AL-MEJIBLI, I. S., ALWAN, J. K., ABD, & HAMED, D. 2020. The effect of gamma value on support vector machine performance with different kernels. International Journal of Electrical and Computer Engineering (IJECE), 10, pp.5497-5506.

CHARLEONNAN, A., FUFAUNG, T., NIYOMWONG, T., & CHOKCHUEYPATTANAKIT, W. 2016. Predictive Analytics for Chronic Kidney Disease. Management and Innovation Technology International Conference (MITicon).

DESAI, M. 2019. Early Detection and Prevention of Chronic Kidney . 5th International Conference On Computing, Communication, Control And Automation (ICCUBEA).

DEVIKA, R., AVILALA, S. V., & SUBRAMANIYASWAMY, V. 2019. Comparative Study of Classifier for Chronic. 3rd International Conference on Computing Methodologies and Communication (ICCMC).

FERNANDO, M. P., CESAR, F., DAVID, N., JOSE, H. O. 2021. Missing the missing values: The ugly duckling of fairness in machine learning. International Journal of Intelligent Systems.

GHOSH, S., DASGUPTA, A., SWETAPADMA, A. 2019. A Study on Support Vector Machine based Linear and Non-Linear Pattern Classification. International Conference on Intelligent Sustainable Systems (ICISS).

HOQUE dan ALJAMAAN. 2021. Impact of Hyperparameter Tuning on Machine Learning Models in Stock Price Forecasting. IEEE Access.

MING, T., YI, Z., & SONGCAN, C. 2003. Improving Support Vector Machine Classifier by Improving Support Vector Machine Classifier by Based on the Best Distance Measurement. Proceedings of the 2003 IEEE International Conference on Intelligent Transportation Systems.

PUJIANTO, U., RAMADHANI, N. A., WIBAWA, A. P. 2018. Support Vector Machine with Purified K-Means Clusters for Chronic Kidney Disease Detection. The 2nd East Indonesia Conference on Computer and Information Technology (EIConCIT), pp.56-60.

QUE, Q., dan BELKIN, M. 2017. Back To The Future: Radial Basis Function Network Revisited. IEEE Transactions On Pattern Analysis and Machine Intelligence.

RAJU, V. N. G., LAKSHMI, K. P., JAIN, V. M., KALIDINDI, A., & PADMA, V. 2020. Study the Influence of Normalization/Transformation process on the Accuracy of Supervised Classification. Third International Conference on Smart Systems and Inventive Technology (ICSSIT).

RUBINI, L. J., PANDIAN, S., & ESWARAN. 2015. Chronic_Kidney_Disease Data Set. [online] Tersedia di: <https://archive.ics.uci.edu/ml/datasets/chronic_kidney_disease> [Diakses 10 Januari 2021]

WU, Z., XIONG, Y., YU, S. X., LIN, D. 2018. Unsupervised Feature Learning via Non-Parametric Instance Discrimination. IEEE/CVF Conference on Computer Vision and Pattern Recognition.

ZHANG, S. 2021. Challenges in KNN Classification. IEEE Transactions on Knowledge and Data Engineering.

Diterbitkan

14-04-2023

Terbitan

Bagian

Ilmu Komputer

Cara Mengutip

Identifikasi Gagal Ginjal Kronis dengan Mengimplementasikan Metode Support Vector Machine beserta K-Nearest Neighbour (SVM-KNN). (2023). Jurnal Teknologi Informasi Dan Ilmu Komputer, 10(2), 301-308. https://doi.org/10.25126/jtiik.20231026059