Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2023
- Goals
- Record
- 767f49c6-46cd-4094-9fef-b5cd9f1e33e6
- Published
- 2026-10-30
Universitas Dian Nuswantoro conducted an activity on the early detection and classification of melanoma skin cancer using the Convolutional Neural Network method with the MobileNet V2 architecture. The activity involved Prabandari, Jasmine Iswarini Dyah; Sari, Christy Atika; Rachmawanto, Eko Hari; Irawan, Candra; Altamer, Bilal R.; and Alkhafaji, Mohamed Ayad as participants. The activity addressed melanoma as a malignant form of cancer affecting the skin and having a particularly high mortality rate. It focused on early detection to increase the level of safety for users.
The activity considered the limitations of diagnosis and skin cancer detection through manual screening and visual inspection. According to the source, this process requires a long time, has high complexity, is subjective, and is prone to errors. To address the classification task, the activity used two classes: benign and malignant. The proposed method applied a Convolutional Neural Network and used the MobileNet V2 architecture to detect melanoma skin cancer.
The proposed method yielded an accuracy of 81.11% for the validation data. The source states that the accuracy results can be improved by using more datasets and increasing the number of layers used. In relation to the supplied goal, this activity concerns early detection and classification of melanoma skin cancer. The activity records UDINUS involvement in applying a Convolutional Neural Network approach to the stated detection task while retaining the stated scope of two melanoma classes.