University of Amikom Yogyakarta

The MIT-BIH Polysomnography Database (SLPDB) is a widely adopted benchmark for the development of automated methods for sleep disorder detection and sleep stage classification. This study presents a Systematic Literature Review of 35 articles that utilize the SLPDB, examining research focus areas, types of physiological signals employed, and the computational approaches applied. Five major methodological categories were identified: Sleep Apnea Detection, Sleep Staging, Signal Processing Enhancement, Multichannel Fusion Methods, and Interpretable Artificial Intelligence, with the first two categories being the most dominant. Four groups of physiological signals—EEG, ECG, respiratory signals, and multichannel data—form the basis for model development, where EEG is predominantly used for sleep staging and ECG for sleep apnea detection. Deep learning approaches, particularly CNNs, LSTMs, and hybrid models, are the most frequently employed techniques. Reported model accuracies range from 78% to over 99%, depending on the signal modality and modeling strategy. Future research should prioritize the development of more interpretable hybrid models and broader clinical validation to enhance reproducibility and implementation readiness.

With the rapid proliferation of JPEG compression in digital communications, the demand for reliable No-Reference (NR) quality assessment has intensified. This paper proposes a content-adaptive blind JPEG image quality assessment technique via multi-domain feature fusion. The proposed framework addresses the limitation of existing metrics, which frequently misidentify complex scene textures as distortion. A comprehensive set of 20 candidate features is extracted, spanning frequency-spectral sparsity, spatial block artifacts, and natural scene statistics. To eliminate variables that bias the model toward image content, a Content-Adaptive Feature Selection (CAFS) mechanism utilizing Recursive Feature Elimination with Cross-Validation (RFE-CV) is deployed. The optimized feature subset is utilized to construct two distinct quality assessment models. The primary model, FuIQA, employs a feed-forward neural network tuned via Bayesian Optimization to capture complex non-linear mappings. In addition, a computationally efficient variant, FuIQA-Lite, is proposed using a linear regression framework to offer a transparent, closed-form mathematical solution ideal for resource-constrained edge devices. Experimental validation across four diverse datasets (Kodak, UCID, USC-SIPI, and Waterloo) confirms that both variants significantly outperform traditional evaluators and modern deep learning baselines such as NIMA. The models demonstrate robust cross-dataset generalizability; when trained on the Waterloo dataset, the non-linear FuIQA achieved an average blind cross-dataset R-squared of 0.9925 and a PLCC of 0.9962, while the linear FuIQA-Lite achieved an SROCC of 0.9808.

Penelitian ini bertujuan untuk menganalisis pengaruh kecerdasan emosional, kecerdasan intelektual, dan kecerdasan spiritual terhadap kinerja karyawan marketing pada PT. Nasmoco Bahana Motor Kota Yogyakarta. Metode penelitian yang digunakan adalah kuantitatif dengan pendekatan survei. Populasi penelitian berjumlah 50 orang karyawan marketing, yang sekaligus diambil sebagai sampel secara sensus. Data dikumpulkan melalui kuesioner dan dianalisis menggunakan regresi linier berganda dengan bantuan SPSS 16. Hasil penelitian menunjukkan bahwa: (1) kecerdasan emosional berpengaruh positif dan signifikan terhadap kinerja karyawan; (2) kecerdasan intelektual berpengaruh positif dan signifikan terhadap kinerja karyawan; (3) kecerdasan spiritual berpengaruh positif dan signifikan terhadap kinerja karyawan. Ketiga variabel secara bersama-sama memberikan kontribusi sebesar 61,0% terhadap variasi kinerja karyawan. Implikasi penelitian ini menekankan pentingnya pengembangan ketiga aspek kecerdasan secara holistik dalam upaya meningkatkan kinerja sumber daya manusia di bidang pemasaran.

The global beauty and personal care products market reached USD 557.24 billion in 2023 and is projected to reach USD 937.13 billion by 2030, growing at a CAGR of 7.7%. This growth is driven by increasing consumer awareness of personal appearance and the proliferation of e-commerce platforms, particularly in the Asia-Pacific region. However, the wide variety of skincare products creates information overload, making it difficult for consumers to identify products that suit their specific needs. Traditional recommendation systems generally rely on a single approach, either content-based filtering, which struggles to capture user preferences, or collaborative filtering, which faces cold-start and data sparsity issues. This study proposes a hybrid ensemble learning approach that integrates three complementary techniques: (1) TF-IDF content-based filtering to analyze product similarities based on brand, category, and product attributes; (2) SVD matrix factorization collaborative filtering to capture latent patterns of user-product interactions through synthetic user data generation; and (3) Random Forest as a meta-learner to intelligently combine the outputs of the two methods. The proposed system was evaluated using a dataset of more than 7,500 skincare products from Sociolla via the Kaggle repository, covering more than 300 brands with detailed product attributes including ratings, reviews, prices, and customer engagement metrics. The hybrid ensemble approach showed strong predictive performance with an R² score of 0.830, explaining 83.0% of the variance in product ratings. The system successfully recommended five products from five different brands (Biyu, True-to-skin, Jacquelle, The-aubree, and Biore) with ensemble scores ranging from 0.872 to 0.996, demonstrating cross-brand recommendation capabilities. Feature importance analysis shows relatively equal contributions from log_wishlist (29.7%), brand_encoded (25.6%), log_reviews (24.1%), and category_encoded (20.6%). These findings indicate that this hybrid approach effectively overcomes the limitations of single-method recommendation systems and can be adapted to various e-commerce product categories, providing a more relevant and personalized shopping experience for consumers.

Dalam menjalankan tugasnya, guru Pendidikan Anak Usia Dini (PAUD) yang menjadi salah satu faktor kunci dalam membentuk dasar perkembangan kognitif, bahasa, sosial-emosional, dan karakter anak, dituntut memiliki kemampuan komunikasi lisan yang efektif. Pelatihan public speaking bagi guru PAUD diadakan untuk menjembatani permasalahan adanya keterbatasan kompetensi public speaking dan kemampuan bercerita (storytelling) yang berdampak pada kurang optimalnya penyampaian pesan edukatif di lingkungan PAUD. Melalui Kerjasama antara Universitas Amikom Yogyakarta dan Yayasan Pendidikan Muamalat Nadhlatul Ulama Daerah Istimewa Yogyakarta (YPMNU DIY), kegiatan pengabdian kepada masyarakat yang melibatkan 60 peserta ini melalui dua sesi utama, yaitu Effective Public Speaking dan Storytelling. Metode pelatihan meliputi penyampaian materi, diskusi terarah, simulasi praktek, umpan balik sejawat, serta refleksi terstruktur. Evaluasi program dilakukan melalui tanggapan, saran, dan testimoni baik dari peserta maupun pengurus YPMNU. Program semacam ini diharapkan berkontribusi pada peningkatan profesionalisme guru PAUD dalam menciptakan proses pembelajaran yang komunikatif, inspiratif, dan bermakna bagi anak usia dini.