IMPLEMENTASI REGRESI LOGISTIK BINER DENGAN REGULARISASI LASSO DALAM MENANGANI MASALAH OVERFITTING (STUDI KASUS: KANKER PAYUDARA WISCONSIN)

Sindi Lestari, NIM.: 21106010036 (2025) IMPLEMENTASI REGRESI LOGISTIK BINER DENGAN REGULARISASI LASSO DALAM MENANGANI MASALAH OVERFITTING (STUDI KASUS: KANKER PAYUDARA WISCONSIN). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (IMPLEMENTASI REGRESI LOGISTIK BINER DENGAN REGULARISASI LASSO DALAM MENANGANI MASALAH OVERFITTING (STUDI KASUS: KANKER PAYUDARA WISCONSIN)) 21106010036_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (6MB) | Preview
	Text (IMPLEMENTASI REGRESI LOGISTIK BINER DENGAN REGULARISASI LASSO DALAM MENANGANI MASALAH OVERFITTING (STUDI KASUS: KANKER PAYUDARA WISCONSIN)) 21106010036_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (8MB) | Request a copy

Abstract

Binary logistic regression is often used to predict variables with two possible outcomes. In application, overfitting can occur when the model overfits the training data and fails to provide accurate predictions on new data. To overcome overfitting, regularization techniques can be used such as the LASSO method which adds a penalty function to the size of the regression coefficients. This study aims to find out how the application of LASSO method helps to reduce overfitting and improve the ability of model to deal with new data. With the LASSO method, some regression parameters will be penalized, so that their coefficient values become smaller or even zero, which makes the model simpler and less complicated. The results show that applying the LASSO method to the binary logistic regression model improves the accuracy and generalization ability of the model. Before the application of the LASSO method, the accuracy on the training data reached 100%, while on the test data was only 96%. After applying LASSO, the accuracy on the test data increased to 98,23%, while the accuracy on the training data decreased to 98,24%, which indicates that the model becomes more balanced and no longer adjusts too much to the training data. The LASSO method also simplifies the model by eliminating irrelevant variables and producing a more efficient model

Share this knowledge with your friends :

Actions (login required)

View Item