Hasil Pencarian  ::  Simpan CSV :: Kembali

"Ground moving radartarget classification is one of the recent research issues that has arisen inan airborne ground moving target indicator (GMTI) scenario. This work presentsa novel technique for classifying individual targets depending on their radarcross section (RCS) values. The RCS feature is evaluated using the Chebyshevpolynomial. The radar captured target usually provides an imbalanced solutionfor classes that have lower numbers of pixels and that have similarcharacteristics. In this classification technique, the Chebyshev polynomial?sfeatures have overcome the problem of confusion between target classes withsimilar characteristics. The Chebyshev polynomial highlights the RCS featureand is able to suppress the jammer signal. Classification has been performed byusing the probability neural network (PNN) model. Finally, the classifier withthe Chebyshev polynomial feature has been tested with an unknown RCS value. Theproposed classification method can be used for classifying targets in a GMTIsystem under the warfield condition."

"Ground moving radar target classification is one of the recent research issues that has arisen in an airborne ground moving target indicator (GMTI) scenario. This work presents a novel technique for classifying individual targets depending on their radar cross section (RCS) values. The RCS feature is evaluated using the Chebyshev polynomial. The radar captured target usually provides an imbalanced solution for classes that have lower numbers of pixels and that have similar characteristics. In this classification technique, the Chebyshev polynomial’s features have overcome the problem of confusion between target classes with similar characteristics. The Chebyshev polynomial highlights the RCS feature and is able to suppress the jammer signal. Classification has been performed by using the probability neural network (PNN) model. Finally, the classifier with the Chebyshev polynomial feature has been tested with an unknown RCS value. The proposed classification method can be used for classifying targets in a GMTI system under the warfield condition."

"Cylindrical uranium dioxide pellets, which are the main components for nuclear fuel elements in light water reactors, should have a high density profile, a uniform shape, and a minimum standard quality for their safe use as a reactor fuel component. The quality of green pellets is conventionally monitored by laboratory measurement of the physical pellet characteristics; however, this conventional classification method shows some drawbacks, such as difficult usage, low accuracy, and high time consumption. In addition, the method does not address the non-linearity and complexity of the relationship between pellet quality variables and pellet quality. This paper presents the development and application of a modified Radial Basis Function neural network (RBF NN) as an automatic classification system for green pellet quality. The weight initialization of the neural networks in this modified RBF NN is calculated through an orthogonal least squared method, and in conjunction with the use of a sigmoid activation function on its output neurons. Experimental data confirm that the developed modified RBF NN shows higher recognition capability when compared with that of the conventional RBF NNs. Further experimental results show that optimizing the quality classification problem space through eigen decomposition method provides a higher recognition rate with up to 98% accuracy."

"ABSTRAKTelah dibuat sebuah alat pencuci tangan otomatis dengan memanfaatkanmikrokontroler AT89S51. Produk-produk sanitari yang ada belum memanfaatkan secaramaksimal proses pengotomatisan. Padahal dengan otomatisasi misalnya pada sistempencuci tangan, kita dapat mengatur volume air dan sabun rata-rata yang dibutuhkanpengguna untuk mencuci tangan sehingga dapat mengurangi penggunaan air dan sabunyang berlebihan. Sistem ini diatur melalui program dengan menggunakan software Reads51. Untuk proses otomasi, digunakan sensor Passive Infra Red (PIR). Sensor inidiletakkan pada saluran keluaran fluida dan blower. Jika sensor PIR pada salurankeluaran fluida mendeteksi adanya tangan, maka saluran keluaran fluida akanmengalirkan air selam 8 detik, sabun selama 6 detik, kemudian air untuk membilasselama 12 detik. Tetapi jika sensor PIR pada blower yang mendeteksi adanya tanganmaka blower sebagai pengering tangan akan menyala selama 12 detik. Denganpendesaian alat ini maka diharapkan akan menjadikan proses pembelajaran otomasidengan lebih mudah dan lebih efektif."

"ABSTRAKAutomatic schema matching is a process to find correspondences among different data attributes from either databases or XML schemas. since there is an inconsistency for naming attributes, the schema matching which is done by humans is the most practical; however, it is time-consuming and incurs great expense. therefore, automatic schema matching process has been extensively studied in the past. Most works still face many challenges such as abbreviation, synonym, hypernym, and structural problems. Some existing works take schema name, instance, data type and schema description as internal resources while other works employ external resources, such as several online dictionaries and ontologies, to increase accuracy for schema matching.In this paper, we address automatic matching problems by employing abbreviation, synonym, and hypernym lists; furthermore, we propose a novel structure similarity algorithm. Finally, we propose to use fuzzy logic, a novel fuzzy scoring algorithm to increase the accuracy of our system. as comparing our systems with existing works on open data; we find that our system outperforms existing works with an f-measure of 90%."

"Universitas XYZ sebagai institusi Perguruan Tinggi Terbuka Jarak Jauh (PTTJJ), senantiasa menjaga kualitas layanannya agar tetap berkualitas. Salah satu layanan yang senantiasa dijaga adalah layanan Bahan Ajar. Layanan Bahan Ajar didukung dengan manajemen stok bahan ajar dari mulai perencanaan dengan melakukan estimasi kebutuhan bahan ajar, gudang bahan ajar untuk menyimpan persediaan bahan ajar dan Student Record System (SRS). Bahan Ajar disiapkan dalam dua program yaitu melalui Sistem Paket Semester (Paket) dan non-paket. Mahasiswa yang mengikuti program nonpaket tidak diwajibkan membayar tagihan biaya bahan ajar. Untuk menjaga kualitas layanan bahan ajar, Universitas XYZ melakukan estimasi kebutuhan bahan ajar. Estimasi dilakukan secara manual dengan menggunakan formula yang berbeda pada setiap tahunnya. Estimasi dilakukan sebelum dan sesudah masa registrasi mata kuliah. Kenyataannya, kebutuhan bahan ajar masih mengalami kekurangan. Hal ini diketahui pada akhir tahun terdapat perbedaan antara hasil estimasi dan realisasi, sehingga tidak sedikit mahasiswa mendapatkan bahan ajar ketika memasuki akhir semester bahkan ketika memasuki awal semester baru. Penelitian ini bertujuan untuk menentukan berapa banyak bahan ajar yang harus disiapkan dengan cara mempelajari profil mahasiswa melalui data history mahasiswa menggunakan teknik classification. Metode yang digunakan Naïve Bayes, Decision Tree dan Support Vector Machine. Evaluasi menggunakan metode cross validation dengan nilai k 2, 3, 5 dan 10. Hasil percobaan menunjukkan bahwa metode Decision Tree memiliki accuracy tertinggi dibanding dengan yang lain.

---

XYZ University as an institution of Distance Learning Higher Education (PTTJJ), always maintains the quality of its services to remain qualified. One service that is always maintained is the Teaching Materials service. Teaching Material Services are supported by the management of teaching material stocks from the start of planning by estimating teaching material requirements, warehouse of teaching materials to store supplies of teaching materials and Student Record System (SRS). Teaching Materials are prepared in two programs, namely through the Semester Package System (Package) and nonpackage. Students who take non-package programs are not required to pay bills for teaching materials. To maintain the quality of teaching material services, XYZ University estimates the need for teaching materials. Estimates are done manually by using a different formula each year. Estimates are made before and after the registration period of the course. In fact, the need for teaching materials is still lacking. This is known at the end of the year there is a difference between the results of estimation and realization, so that not a few students get teaching materials when entering the end of the semester even when entering the beginning of the new semester. This study aims to determine how much teaching material must be prepared by studying student profiles through student history data using classification techniques. The method used is Naïve Bayes, Decision Tree and Support Vector Machine. The evaluation uses the cross validation method with values k 2, 3, 5 and 10. The experimental results show that the Decision Tree method has the highest accuracy compared to the others."

"ABSTRAKDewasa ini, telemarketing banyak digunakan oleh perusahaan sebagai sarana untuk meningkatkan penjualan. PT XYZ adalah perusahaan yang bergerak dalam bidang e-commerce classified ads customer-to-customer (C2C). PT XYZ melakukan telemarketing sejak bulan Juni 2016. Akan tetapi, rata-rata konversi telemarketing per minggu hanya 5,1 persen. Penelitian ini menggunakan metodologi CRISP-DM. Penelitian ini menggunakan Customer Lifetime Value (LTV) sebagai pembentuk fitur. Teknik Synthetic Minority Over Sampling (SMOTE) untuk permasalahan kelas tidak berimbang memberikan hasil yang lebih baik dari teknik undersampling dalam penelitian ini. Algoritma klasifikasi yang diuji adalah bayesian network, decision tree, random forest, support vector machince, neural network, bagging neural network, deep neural network, adaboost deep neural network, convolutional neural network, dan extreme gradient boosting tree. Evaluasi terhadap model yang dihasilkan menitikberatkan pada analisis cost-benefit, dan analisis gain chart. Analisis cost-benefit terhadap gain chart menunjukkan keuntungan terbaik ada pada desil 20 persen populasi. Analisis cost-benefit menunjukkan bahwa algoritma adaboost deep neural network dengan data SMOTE mendapatkan hasil terbaik. Akan tetapi model random forest SMOTE dapat mengklasifikasikan 73,13 persen pelanggan fitur premium dalam 20 persen populasi, serta memiliki konsistensi terbaik dalam gain chart. Total potensi keuntungan adalah 11,5 kali lipat dibandingkan tanpa menggunakan model. Dalam penelitian ini ditemukan juga bahwa nilai Fβ-score dapat digunakan untuk mengukur nilai cost-benefit dari model klasifikasi yang dihasilkan.

---

ABSTRACTMany companies have widely used telemarketing to increase their sales. XYZ Company focusing their business on classified ads e-commerce platform. XYZ Company have been using telemarketing since June 2016, but weekly conversion rate from telemarketing is only 5,10 percent. CRISP-DM methodology serves as guideline principle in this research. Customer Lifetime Value (LTV) was used to guide feature extraction. In this research, Synthetic Minority Over Sampling (SMOTE) gives better result than undersampling to tackle class imbalance problem. Algorithm tested are bayesian network, decision tree, random forest, support vector machince, neural network, bagging neural network, deep neural network, adaboost deep neural network, convolutional neural network, and extreme gradient boosting tree. This research emphasizes on cost benefit and gain analysis to evaluate the result. Cost benefit analysis on gain chart shows the best benefit achieved on 20 percent of the population. AdaBoost Deep Neural Network with SMOTE data gives best result on cost-befenit analysis. However, Random Forest and SMOTE data could recognize 73.13 percent of target only in 20 percent of the population, and it gives the best consistency across 20 to 50 percent of the population. Total potential benefit is 11.5 times higher than without using model. We also found that Fβ-score values were correlated with cost-benefit values from classification model."

"The performance and usability of the input device play an important role in providing better experience for the user. The touchpad is commonly known as a pointing device and is a predominant pointing technology for notebook computers. However, comparative evaluations have established that touchpad performance is poor in comparison with a mouse. The best setting of touchpad is also remaining unknown. Furthermore, there is no research that study about the velocity pattern in touchpad. To solve this drawback, this research attempts to implement Fitts' Law method, merely focused on touchpad. In the design of experiment, touchpad size and position filter are added as new independent variables, along with Control Display Gain, Distance, Width, and Angle, as the wellknown variables in Fitts' Law researches. Two sizes of touchpad are prepared which consist of large (100*60) and small (65*36) sizes. In addition, position filter is set at 2 different levels: 30 and 50, moreover gain setting is set at 3 different levels of fixed gain: 0.5, 1, and 2. For the Fitts' Law Program, 3 different levels of distance (100, 300, and 500 pixel), 3 different levels of target width (10, 40, and 70 pixel), and 8 directions (0, 45, 90, 135, 180, 225, 270, and 315) are applied. Moreover, the dependent variables that are being studied are movement time, error count, movement count, target re-entry count, and peak velocity. In this experiment, 20 participants are recruited and ANOVA Split Plot is used as the method. In total, each participant performed 2592 trial movements (2 touchpad size × 3 position filter × 3 control display gain x 3 distance × 3 target size × 8 moving direction × 3 repetitions). As for the results, touchpad size significantly affects movement time, error count, movement count, and re-entry count. Position filter also significantly affects the re-entry count. The best setting acquired from result shows that filter 50 and gain 2 are better implemented for primary movement, and filter 30 and gain 0.5 are better applied in secondary movement. The result also shows that there is difference in angle for touchpad performance and mouse. The different behavior for touchpad user also differs in touchpad performance indicator. Moreover, clutching behavior on touchpad user makes touchpad velocity graph to be modeled into several primary movement. Furthermore, strong interaction between distance and gain influences Fitts' Law equation to be modified."

"ABSTRAKSaat ini, Generasi Z adalah generasi dengan kepribadian dan karakteristik paling kompleks dan kritis sepanjang masa. Perbedaan karakter dari generasi terdahulunya sangat mempengaruhi perilaku pembelian dari Generasi Z. Di tahun 2020 mendatang, diperkirakan bahwa anggota dari Generasi Z akan menjadi kelompok konsumen terbesar di dunia, melebihi berbagai kelompok generasi sebelumnya. Oleh karena itu, kekuatan dan daya beli dari generasi ini tidak dapat disepelekan. Walau kini anggotanya berusia maksimal 22 tahun, pengaruh mereka di lingkungan sosial dan keluarga terbilang signifikan. Kekuatan dari Generasi Z diperkuat dari mudahnya akses informasi dan fasihnya mereka untuk berkomunikasi di dalam dunia teknologi dan digital. Sebagai generasi pertama yang lahir sebagai digital natives, mereka hidup tidak hanya di dunia nyata melainkan juga di dunia digital. Berdasarkan kebutuhan dan perilaku yang berbeda, brand perlu mempertimbangkan ulang taktik dan strategi pendekatan kepada target baru ini. Tentunya, strategi pemasaran digital adalah pendekatan yang perlu ditekankan. Melihat potensi pembelian yang besar dari generasi ini, brand yang visioner akan mulai membidik Generasi Z sebagai target utama mereka, dengan pendekatan bauran pemasaran yang sesuai dengan karakteristiknya.

---

ABSTRACTNowadays, Generation Z is the most complex and critical generation from all times. These character and behavior differences from its preceding generations affected their buying decision process. I tis estimated that by 2020, Generation Zs will come to be the largest group of consumers globally. Judging by the statement given, this younger genereration should not be looked down by brand marketer. With the oldest member barely even 22 years old, their influence both in social and family environment can be considered as significant. Their favorable state of life is also supported by information accessibility and their fluency in the digital technology. They are the first ever digital natives born generation, hence familiar living both in reality and varied social networks. Their specific behavior and characteristics require brands to reconsider their current approach and strategy to engage with this new audience. Digital marketing is a one definite strategy to be focused on while targeting to Generation Z. Seeing the potential of thie generation, a visionary brand will start targeting this generation as their main audience, with a tailored marketing mix approach to meet their expectations."

"The NTIS classification system has simpler notation than DDC. It does not recognize supporting table so that it can determine the notation faster. The number of the main class in NTIS classification system is 39 while DDC is 10. NTIS is most suitable for special libraries in the field of technology such as the Agency for the Assesment and Application of Technology (BPPT) since the system has a more specific technology subject. DDC is also effective for collection data exchange since 2010, referring that generally libraries in Indonesia has not recognized NTIS. Both systems actually have its advantages and disadvantages. In determining notations, both NTIS and DDC use the same initial step that is to determine the collection subject. NTIS is faster especially in handling technology subject. It is also more specific in referring technology subject. The number of the main class in NTIS is 39 while DDC is 10. Both systems have index. NTIS does not have supporting table while DDC has. NTIS uses a simpler notation because it uses only 2 digits. According to a key informant, the NTIS classification system does not recognize supporting table. In terms of notation search, NTIS's scheme is faster because it uses limited classification numbers. Index is mostly used for determining classification notation. Index in both systems is a clue represented in a systematically arranged letters. In NTIS, it can be figured out that subjects on technology is more specific yet in some certain categories is not as detail as DDC."