Dinda Cantika Putri, . (2026) PENERAPAN CONVOLUTIONAL NEURAL NETWORK DENGAN MOBILENETV3 UNTUK DETEKSI JENIS KULIT WAJAH SEBAGAI REKOMENDASI KANDUNGAN SKINCARE BERBASIS MOBILE. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (225kB) |
|
Text
AWAL.pdf Download (862kB) |
|
Text
BAB I.pdf Restricted to Repository UPNVJ Only Download (279kB) |
|
Text
BAB II.pdf Restricted to Repository UPNVJ Only Download (972kB) |
|
Text
BAB III.pdf Restricted to Repository UPNVJ Only Download (865kB) |
|
Text
BAB IV.pdf Restricted to Repository UPNVJ Only Download (2MB) |
|
Text
BAB V.pdf Download (256kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (239kB) |
|
Text
RIWAYAT HIDUP.pdf Restricted to Repository staff only Download (196kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (3MB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (33MB) |
|
Text
ARTIKEL KTI.pdf Restricted to Repository staff only Download (718kB) |
Abstract
The demand for daily skincare continues to rise alongside urban lifestyles and intensive academic activity, yet most individuals, including university students, still do not fully understand their facial skin type or which ingredients suit their skin condition, risking incorrect product choices that can trigger irritation and allergies. This research develops a mobile application called Jelita that automatically detects facial skin type and provides suitable skincare recommendations. The methods applied include a Convolutional Neural Network with a transfer-learning-based MobileNetV3-Large architecture strengthened by a Convolutional Block Attention Module (CBAM) to classify five facial skin types, namely normal, dry, oily, acne-prone, and combination, along with Content-Based Filtering using TF-IDF weighting and cosine similarity to recommend skincare products based on data extracted from the Female Daily platform. The application was built using Flutter, FastAPI, and Supabase, following an Agile development approach. Testing results show that the CNN MobileNetV3 model achieved a test accuracy of 91.63%, with average precision, recall, and f1-score of 90%. Black box testing on 22 functional scenarios showed all features performing as expected, while the User Acceptance Test (UAT) involving seven respondents achieved a feasibility percentage of 91.43%, categorized as very good, proving that Jelita is suitable as a tool for more objective and personalized skin type detection and skincare recommendation.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil: 2210511054] [Pembimbing 1: Neny Rosmawarni] [Pembimbing 2: Radinal Setyadinsa] [Penguji 1: Muhammad Adrezo] [Penguji 2: Anis Fitri Nur Masruriyah] |
| Uncontrolled Keywords: | Content-Based Filtering, Convolutional Neural Network, Facial Skin Type, MobileNetV3, Skincare |
| Subjects: | Q Science > QA Mathematics Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | DINDA CANTIKA PUTRI |
| Date Deposited: | 29 Aug 2026 14:31 |
| Last Modified: | 29 Aug 2026 14:31 |
| URI: | http://repository.upnvj.ac.id/id/eprint/51997 |
Actions (login required)
 |
View Item |