Abstract.One of new student admission pathways at Universitas Negeri Surabaya (Unesa) is through the Indonesian National Public University Admission. This path is quite favourable for academic or vocational high school students who want to study at Unesa so that the number of participants for this selection program can reach up to ten thousand people every year. The large number of applicants makes the selection process more complex. Meanwhile, Unesa still uses the concept of weighting criteria in determining the result. One of the constraints in weighting process is the absence of optimal pattern or weight. This paper discusses a supervised learning approach to make determining process pattern which replaces manual weighting criteria. The supervised learning method used in this research was the Neural Network with multilayer perceptron. This research showed sufficient result which can be seen from the high accuracy rate (89.56%). The accuracy rate is enough to decide which participants who pass or fail the national admission. This system can be used as a prediction for the following years.
Mesin oven untuk UKM saat ini masih menggunakan teknologi yang opresionalnya secara manual. Agar lebih efisien maka dan merupakan tuntutan dari era Industri 4.0 maka diperlukan penggunaan teknologi untuk mesin-mesin yang digunakan oleh UKM di Indonesia. Tujuan dari peelitian ini adalah merancang bangun mesin oven pengering cerdas berbasis internet of things (IoT) ini, sumber energi panasnya didesain bisa menggunakan panas dari kompor LPG, minyak tanah, kayu bakar, maupun batubara briket. Selanjutnya panas dialirkan melalui pipa besi yang berfungsi sebagai jalur keluarnya asap ke cerobong pembuangan dan juga berfungsi penyimpan energi panas, sehingga terjadi proses perpindahan panas dari pipa penyimpan panas ke udara di ruangan tersebut. Udara dalam ruangan tersebut akan terkondisikan. Sistem Internet of Things (IoT) berfungsi sebagai pemantauan (monitoring) proses pengeringan berjalan dengan baik. Berdasarkan eksperimen yang telah dilakukan sistem telah dapat bekerja dengan baik melalui offline dan online. Sistem kontrol dapat bekerja dengan menggunakan sistem arduino maupun melalui online dengan berbasis IoT. Setelah dilakukan pengujian dengan memberikan perlakuan suhu yang berbeda-beda menunjukkan sistem bekerja dengan baik.
Abstrak— Seiring dengan perkembangan teknologi informasi, kebutuhan dalam pencarian informasi menjadi hal yang penting. Jika pencarian informasi selama ini dilakukan pada data berjenis teks, maka pada perkembangan teknologi saat ini, memungkinkan adanya pencarian informasi dalam bentuk citra digital. Hal tersebut terjadi karena adanya peningkatan jumlah pustaka digital dalam bentuk citra. Sebuah metode pengembalian citra menjadi komponen utama untuk memecahkan masalah tersebut. CBIR merupakan sistem pengembalian citra yang akan membantu dalam proses pencarian citra dengan memanfaatkan fitur-fiturnya. Penggunaan ekstraksif fitur yang tepat diperlukan untuk mendapatkan fitur tersebut. Pemilihan ekstraksi ftur akan sangat memengaruhi hasil dari CBIR. Salah satu metode yang dapat melakukan ekstraksi fitur pada citra adalah CNN. Metode yang masih dalam satu jenis dalam deep learning ini mampu mempelajari fitur citra untuk dimanfaatkan ke dala bidang visi komputer. Karena itu, CNN menjadi perhatian menarik dalam penelitian ini untuk melakukan CBIR. Penggunaan filter Gabor yang mampu mendapatkan tekstur citra dengan baik juga akan diimplementasikan sebagai filter pada lapisan konvolusi CNN. Dengan menggunakan CNN dan filter gabor, penelitian ini mampu mendapatkan nilai mAP sebesar 0,895 terhadap data uji dengan dataset GHIM10k. Penelitian ini juga membandingkan beberapa metode pengukuran jarak untuk mendapatkan sistem CBIR terbaik. Kata Kunci— Content Based Image Retrieval; Convolutional Neural Networks; pengukuran jarak; filter Gabor; visi komputer.
Praktik Industri (PI) is one of the mandatory courses that must be taken by students of all majors within the Faculty of Engineering, State University of Surabaya. Through this course, students are required to carry out learning in institutions/companies according to the field of science with the aim of knowing and practicing what they have learned in real practice in the company. The main problem in implementing PI is that all administrative processes are not integrated so that it is difficult for the parties involved in managing the PI. Some examples of problems related to PI include student information and industrial premises, activities while in industry, students who have not done PI, PI guidance notes, notes related to PI exams, Student Information and Advisory Lecturers, administrative data related to PI, Feedback from Industry and others. -other. Based on these problems, PI activities require an PI management information system that is integrated with the Integrated Academic Information System (SIAKADU) in the Department of Informatics, Faculty of Engineering Unesa. This PI management information system is built on a web-based basis using open source technology and is integrated with data on SIAKADU.
Abstract. The laboratory is a facility that provides all kinds of equipment necessary for scientific activities. As one of the majors in the Faculty of Engineering, Universitas Negeri Surabaya, Department of Informatics Engineering (JTIF) also has several laboratories to support teaching and learning process for both students and teachers. Lots of equipment stored in the laboratory to support learning in the Department of Informatics. Because many equipment's in the laboratory, it is necessary to record inventory. The records that have been done so far are still manual using MS word or MS excel. One of the weaknesses of manual system is the possibility of missing notebooks, so lab work inventory evidence is missing. Of course, this will reduce the quality value of a laboratory and of course this is very ineffective. In this paper, we propose the application of Quick Response (QR) codes to conduct laboratory inventory. This application use framework Bootstrap which supports responsive web technology. By using the framework, the application can be accessed by mobile phone. This framework can respond to user behavior and environment based on platform size, as well as monitor screen orientation. The outcome is expected to facilitate the process of laboratory inventory in the Department of Informatics Engineering UNESA.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.