Hyperparameter Tuning Model Artificial Neural Network untuk Klasfikasi Penyakit Breast Cancer

M Nejatullah Sidqi (Universitas PGRI Silampari)
Ahmad Marsehan (Universitas PGRI Silampari)
Rio Rio (Universitas PGRI Silampari)
Rani Okta Felani (Universitas PGRI Silampari)

Jurnal Teknik Industri Terintegrasi (JUTIN)
Vol. 7 No. 3 (2024): July

Cancer is a deadly disease that is often experienced unexpectedly. Cancer does not only attack adults and the elderly, even newborn babies can get cancer. Breast cancer is the number one cancer disease in Indonesia and is one of the main causes of death due to cancer. According to Globocan 2020, the number of new cases of breast cancer in Indonesia increased from 396,914 new cancer cases to 68,858 cases (16.6%). There was a trial for early diagnosis of breast cancer carried out on the Breast Cancer dataset, totaling 569 patients with target categories namely Malignant (malignant cancer) and Benign (benign cancer). This diagnosis aims to carry out classification using data mining to extract new information and knowledge. The method applied uses the Artificial Neural Network (ANN) algorithm, using parameters, 1 layer and 250 to 500 epochs. Meanwhile, the activation functions used are ReLU and Sigmoid activation. Then the best classification results were produced by 250 epochs with an accuracy of 98.24%.