Hasil Pencarian  ::  Simpan CSV :: Kembali

"ABSTRAK<>br>Dalam persaingan yang ketat di dunia industri, khususnya dalam industri telekomunikasi semakin ketat. Oleh karena itu, masing-masing operator telekomunikasi berlomba-lomba untuk memberikan layanan terbaik agar dapat menarik konsumen yang lebih banyak. Namun, sebelum membuat suatu layanan telekomunikasi, haruslah ada analisis pasar khususnya pelanggan, agar dapat diketahui layanan seperti apa yang dibutuhkan dan diinginkan oleh pelanggan tersebut. Dengan perkembangan sistem informasi, analisis menggunakan data mining menjadi pilihan. Dalam penelitian ini dibuat suatu model KDD Knowledge Discovery Data dengan menggunakan algoritma sistem rekomendasi yang dikombinasikan dengan social network analysis untuk memprediksi suatu layanan apakah diterima oleh pelanggan atau tidak. Dalam penelitian ini terdapat empat macam model sistem rekomendasi, yaitu collaborative filtering, content based filtering, hybrid filtering dan stochastic gradient descent. Kemudian hybrid filtering dan stochastic gradient descent dikombinasikan dengan hasil analisis dari jaringan sosial yang berupa nilai-nilai pagerank, eigenvector, modularity, degree dan sebagainya. Hasil dari kombinasi hybrid filtering dengan hasil analisis jaringan sosial tidak terlalu signifikan dan cenderung tetap, sedangkan hasil dari kombinasi stochastic gradient descent dengan hasil dari analisa jaringan sosial dapat menurunkan nilai RMSE dan MAE sebesar 0,001 sampai 0,010 dan 0,011 sampai 0,013.

---

ABSTRACT<>br>In the fierce competition in the industrial world, especially in the telecommunications industry is getting tighter. Therefore, each telecommunicationoperator will compete to provide the best service to attract more consumers. However, before performing a telecommunication service, there must be a market analysis, especially customers, to know what kind of service is needed and desired by the customer. With the development of information systems, analysis using data mining becomes an option. In this research will be made a model of KDD Knowledge Discovery Data using a recommendation system algorithm combined with social network analysis to predict a service whether received by customers or not. In this research, there are four models of the recommendation system they are collaborative filtering, content based filtering, hybrid filtering and stochastic gradient descent. Then hybrid filtering and stochastic gradient descent combined with the results of analysis of social networks in the form of PageRank values, eigenvector, modularity, degree and so forth. The result of hybrid filtering combination with social network analysis result is not very significant and tends to remain, while the result of stochastic gradient descent combination with result of social network analysis can decrease RMSE and MAE value about 0.001 to 0.010 and 0.011 to 0.013."

"[ABSTRAKPesatnya perkembangan Deep Learning akhir-akhir ini juga menyentuh ASRberbasis HMM, sehingga memunculkan teknik hibrid HMM-ANN. Salah satuteknik Deep Learning yang cukup menjanjikan adalah penggunaan arsitekturCNN. CNN yang memiliki kemampuan mendeteksi local correlation sesuaiuntuk digunakan pada data spectrum suara. Spectrogram memiliki karakteristiklocal correlation yang nampak secara visual. Penelitian ini adalah eksperimenpenggunaan spectrogram sebagai fitur untuk HMM-CNN untuk meningkatkankinerja ASR berbasis HMM. Penelitian menyimpulkan spectogram dapatdigunakan sebagai fitur untuk HMM-CNN untuk meningkatkan kinerja ASRberbasis HMM.

---

ABSTRACTThe latest surge in Deep Learning affecting HMM based ASR, which give birth tohybrid HMM-ANN technique. One of the promising Deep Learning technique isthe implementation of CNN architecture. The ability of CNN to detect localcorrelation make it suitable to be used for speech spectral data. Spectrogram as aspeech spectral data has local correlation characteristic which is visuallyobservable. This research is an experiment to use spectrogram as a feature forHMM-CNN to add to the performance of HMM based ASR. This research foundthat spectrogram is indeed can be used as a feature for CNN to add to theperformance of HMM based ASR., The latest surge in Deep Learning affecting HMM based ASR, which give birth tohybrid HMM-ANN technique. One of the promising Deep Learning technique isthe implementation of CNN architecture. The ability of CNN to detect localcorrelation make it suitable to be used for speech spectral data. Spectrogram as aspeech spectral data has local correlation characteristic which is visuallyobservable. This research is an experiment to use spectrogram as a feature forHMM-CNN to add to the performance of HMM based ASR. This research foundthat spectrogram is indeed can be used as a feature for CNN to add to theperformance of HMM based ASR.]"

"[ABSTRAKBanyaknya jejaring sosial yang bermunculan. Salah satu jejaring sosial yang marak digunakan adalah twitter. Kegiatan promosi produk melalui twitter sudah mulai digunakan PT XL Axiata Tbk (XL) sejak tahun 2009 melalui akun @XL123. Penggunaan twitter oleh perusahaan telekomunikasi di Indonesia masih dalam tahap penjualan dan promosi.Namun demikian, analisis yang dilakukan hanya terbatas pada perhitungan jumlah retweet, komentar, dan follower. Analisis belum melihat sejauh mana makna komentar dari pelanggan maupun masyarakat. Hal ini akan mempengaruhi keputusan membeli masyarakat jika komentar yang diberikan negatif, dan sebaliknya komentar positif akan meningkatkan citra perusahaan di mata stakeholder. Hal ini dapat dilihat dari fakta bahwa rating yang diperoleh XL Axiata tidak sesuai dengan ekspektasi brand tersebut, yaitu rating 3 dari 10. Sedangkan ekspektasi yang diharapkan berdasarkan analisa perbandingan jumlah follower dan following, semestinya XL Axiata memiliki reputasi yang bagus yaitu 7-8.Penelitian ini melakukan perhitungan reputasi dari produk XL Axiata, dan membandingkannya dengan produk Telkomsel dan Indosat. Selanjutnya dilakukan beberapa teknik ekstrak data, analisis sentimen, serta membandingkan tiga algoritma klasifikasi: Naïve Bayes, Support Vector Machine, dan Decision Tree. Tahap evaluasi performansi menggunakan precision, recall, f-measure, dan kurva ROC (AUC). Hasil menunjukkan bahwa model yang dibentuk oleh SVM memberikan performansi yang lebih baik untuk selanjutnya digunakan untuk melakukan perhitungan Net Brand Reputation. Perhitungan NBR dilakukan di produk 3G, 4G, Voice, SMS, dan Internet (data). Berdasarkan perbandingan kelima produk ini, XL Axiata memperoleh rata-rata nilai reputasi yang lebih di dibandingkan Telkomsel dan Indosat yaitu sebesar 24.5%, sedangkan Telkomsel hanya memperoleh 13.2% dan Indosat 19.3%.

---

ABSTRACTThe internet in Indonesia has grown rapidly, it proved by many social media comes up. One of famous social media is twitter. Campaign product using twitter had been used by XL Axiata since 2009 through account @XL123. Unfortunately, the using of twitter in Indonesia telecommunication company still in the stage of sales and promotions.However, the analysis only calculated number of retweets, comments, dan followers. Analyzes haven?t seen what is the meaning of those comments, whether be positive or negative for XL brand products. Negative comments giving influence to society buying decision, while positive comments create good reputation to stakeholders. This is showned by a fact that the rating obtained XL Axiata does not correspond to the brand?s expectation, ie rating 3 out of 10. While expectation based on comparative analysis of number of followers and following, XL Axiata should have a good reputation in rate 7-8.This study not only calculating XL product but also Telkomsel and Indosat for comparative analysis. Hereafter, we extracted features, algorithms and the classification schemes. Evaluation phase using precision, recall, f-measure and ROC curve (AUC). The sentiments are classified and compared using three different algorithms: Naïve Bayes, Support Vector Machine, and Decision Tree classifier method. The result shows model built by SVM is the best result. Using this model, we measure Net Brand Reputation in 5 products which are 3G, 4G, Voice, SMS, and Internet (data). The experiments showned XL Axiata has the highest reputation score rather than Telkomsel and Indosat with average NBR score 24,5%, while Telkomsel only 13.2% and Indosat 19.3%.;The internet in Indonesia has grown rapidly, it proved by many social media comes up. One of famous social media is twitter. Campaign product using twitter had been used by XL Axiata since 2009 through account @XL123. Unfortunately, the using of twitter in Indonesia telecommunication company still in the stage of sales and promotions.However, the analysis only calculated number of retweets, comments, dan followers. Analyzes haven?t seen what is the meaning of those comments, whether be positive or negative for XL brand products. Negative comments giving influence to society buying decision, while positive comments create good reputation to stakeholders. This is showned by a fact that the rating obtained XL Axiata does not correspond to the brand?s expectation, ie rating 3 out of 10. While expectation based on comparative analysis of number of followers and following, XL Axiata should have a good reputation in rate 7-8.This study not only calculating XL product but also Telkomsel and Indosat for comparative analysis. Hereafter, we extracted features, algorithms and the classification schemes. Evaluation phase using precision, recall, f-measure and ROC curve (AUC). The sentiments are classified and compared using three different algorithms: Naïve Bayes, Support Vector Machine, and Decision Tree classifier method. The result shows model built by SVM is the best result. Using this model, we measure Net Brand Reputation in 5 products which are 3G, 4G, Voice, SMS, and Internet (data). The experiments showned XL Axiata has the highest reputation score rather than Telkomsel and Indosat with average NBR score 24,5%, while Telkomsel only 13.2% and Indosat 19.3%., The internet in Indonesia has grown rapidly, it proved by many social media comes up. One of famous social media is twitter. Campaign product using twitter had been used by XL Axiata since 2009 through account @XL123. Unfortunately, the using of twitter in Indonesia telecommunication company still in the stage of sales and promotions.However, the analysis only calculated number of retweets, comments, dan followers. Analyzes haven’t seen what is the meaning of those comments, whether be positive or negative for XL brand products. Negative comments giving influence to society buying decision, while positive comments create good reputation to stakeholders. This is showned by a fact that the rating obtained XL Axiata does not correspond to the brand’s expectation, ie rating 3 out of 10. While expectation based on comparative analysis of number of followers and following, XL Axiata should have a good reputation in rate 7-8.This study not only calculating XL product but also Telkomsel and Indosat for comparative analysis. Hereafter, we extracted features, algorithms and the classification schemes. Evaluation phase using precision, recall, f-measure and ROC curve (AUC). The sentiments are classified and compared using three different algorithms: Naïve Bayes, Support Vector Machine, and Decision Tree classifier method. The result shows model built by SVM is the best result. Using this model, we measure Net Brand Reputation in 5 products which are 3G, 4G, Voice, SMS, and Internet (data). The experiments showned XL Axiata has the highest reputation score rather than Telkomsel and Indosat with average NBR score 24,5%, while Telkomsel only 13.2% and Indosat 19.3%.]"

"Pasar modal merupakan salah satu penopang perekonomian di Indonesia dikarenakan pasar modal adalah salah satu sarana penambahan modal bagi perusahaan dalam melaksanakan operasional bisnisnya. Selain itu, pasar modal juga dapat memberikan keuntungan bagi perorangan atau badan usaha investor , yang dapat digunakan untuk biaya hidup di masa kini dan masa depan. Saham adalah salah satu instrumen yang diperjualbelikan di pasar modal yang dapat memberikan keuntungan hingga 20-25 per tahun, namun juga memiliki resiko yang lebih tinggi dibandingkan instrumen lainnya, dikarenakan fluktuasi harga saham yang tinggi. Investor harus berhati-hati dalam memilih waktu yang tepat dalam melakukan transaksi jual beli saham, jika tidak ingin mendapatkan kerugian yang besar. Salah satu teknik dalam menentukan waktu transaksi saham adalah berdasarkan teori Efficient Market Hypothesis EMH yang menyatakan bahwa harga saham merefleksikan informasi yang dipublikasikan terkait saham tersebut. EMH berkaitan erat dengan behavioral economics yang menyatakan bahwa keputusan investor dipengaruhi oleh persepsi atau emosi sentimen dari informasi yang diterimanya. Informasi terkait suatu saham dapat dibaca di media online dalam hal ini artikel berita yang diterbitkan oleh berbagai situs berita di Indonesia. Berdasarkan EMH dan behavioral economis tersebut, penelitian ini bertujuan untuk melihat apakah sentimen media online dapat mengklasifikasikan pergerakan harga saham di Indonesia, dengan ruang lingkup media online yang diteliti adalah www.bisnis.com. Saham yang diteliti adalah saham PT. Bank Central Asia, Tbk BBCA. Metode penelitian yang dilakukan adalah penelitian kuantitatif dengan metodologi eksperimental menggunakan pendekatan analisis sentimen. Proses ekstraksi fitur dari artikel berita bisnis.com dilakukan dengan merepresentasikan artikel berita dalam bentuk vektor menggunakan term frequency tf, term frequency inverse document frequency tf-idf, dan doc2vec. Algoritma klasifikasi yang digunakan adalah naive bayes, k-nearest neighbor KNN, dan support vector machine SVM. Setelah mendapatkan hasil klasifikasi sentimen dari berita, penelitian dilanjutkan dengan membandingkan hasil klasifikasi sentimen tersebut dengan harga aktual saham untuk hari itu sampai dengan lima hari ke depan. Hasil penelitian menunjukkan bahwa representasi vektor menggunakan tf-idf menunjukkan hasil yang lebih baik dengan akurasi klasifikasi SVM sebesar 78,66 . Sementara perbandingan akurasi hasil klasifikasi sentimen berita dengan harga aktual pergerakan saham didapatkan sebesar 63,11 . Hasil ini didapatkan untuk kondisi sentimen berita dan harga saham di hari yang sama. Hal ini menunjukkan bahwa para investor dapat mempertimbangkan sentimen berita terkait saham tertentu di situs www.bisnis.com di hari yang sama untuk melakukan transaksi jual beli saham.

---

Capital market has been acted as one of the pillar for economy growth in Indonesia, because capital market is one of additional capital for the company in carrying out its business operations. In addition, capital markets can also provide benefits to individuals or business groups investors , which its profit can be used for the cost of living in the present and the future. Stock Equity is one of the instruments traded in capital markets that can provide gains up to 20 25 per year, but also have a higher risk than other instruments, due to fluctuations in stock prices are high. Investors should be cautious in choosing the right time to make buying or selling shares, if they do not want to get a big loss.

One of technique in determining time for transaction is Efficient Market Hypothesis EMH , which states that stock price movement reflects related stock rsquo s information published. EMH has correlation with behavioral economics which states that investor decision can be influenced by perception or emotion sentiment of the information which he she read. Some information related to specific stock can be read in online media, such as news article from websites in Indonesia. Based on EMH and behavioral economis, this study aims to see whether the sentiment of online media can classify stock price movements, with the scope of online media that will be studied is www.bisnis.com. Stock price which is studied is belong to PT.Bank Central Asia BBCA.JK.

The research method is quantitative research with experimental methodology using sentiment analysis approach. Feature extraction is done by representing news articles in form of vector using term frequency tf, term frequency inverse document frequency tf df, and doc2vec. Classification algorithm used are naive bayes, k nearest neighbors, dan support vector machine SVM. After getting classification result from news sentiment analysis, this study continue by comparing those classifications with actual price of stock that day until five days forward.

The result shows that vector representation using tf idf is better with classification accuracy 78,66. Meanwhile, comparation between news sentiments classification with actual price stock shows 63,11 within the same day. This result means that news article sentiment related to spesific stock from www.bisnis.com can be used as reference for investors in determining their participation in stocks purchase transaction."

"ABSTRAKDi tengah persaingan e-commerce di Indonesia yang semakin ketat, membuat perusahaan e-commerce dituntut dapat bersaing dalam memberikan nilai tambah layanan bagi pelanggannya agar dapat meningkatkan jumlah pelanggan dan disertai dengan angka pemesanan yang meningkat juga. Hal ini juga berlaku bagi e-commerce dalam industri pariwisata, salah satunya Triptrus. Salah satu tantangan yang dihadapi Triptrus adalah untuk meningkatkan angka conversion rate. Salah satu cara peningkatan angka conversion rate adalah pemberian fitur rekomendasi produk. Penelitian ini bertujuan untuk mencari metode yang dapat memberikan rekomendasi terbaik yang pada akhirnya bertujuan agar dapat meningkatkan angka conversion rate dari Triptrus yang masih rendah. Pada penelitian ini dilakukan pencarian metode rekomendasi yang terbaik disesuaikan dengan data internal yang dimiliki Triptrus. Penelitian ini bermula dari pengumpulan data internal untuk kemudian dibangum model rekomendasi menggunakan beberapa metode. Metode yang digunakan dalam penelitian ini diantaranya collaborative filtering, content based filtering, Hybrid Filtering dan stochastic gradient descent. Berdasarkan hasil penelitian, metode collaborative filtering, content based, dan hybrid kurang mampu memberikan rekomendasi yang cukup baik terhadap data Triptrus. Hasil terbaik dari ketiga metode ini didapatkan metode hybrid dengan nilai error RMSE 0.71. Di sisi lain algoritma stochastic gradient descent dapat memberikan rekomendasi paling baik dan memberikan ratio error RMSE paling kecil yaitu 0.11. Hasil penelitian ini adalah model rekomendasi produk yang dapat memberikan rekomendasi terbaik berdasarkan data internal Triptrus yaitu model yang dihasilkan menggunakan metode stochastic gradient descent.

---

ABSTRACTIndonesian e commerce markets are getting more tight. This condition forces e commerce companies to provide value added services for their customer in order to increase numbers of customer, which also followed by increasing number of purchases. This also happened in tourism e commerce company, Triptrus. The challenge that triptrus faces is how to increase their conversion rates. One way to increase the number of conversion rate is the provision of a product recommendation feature. The purpose of this research is to find the best recommendation method that can improve conversion rate in Triptrus. In this research we looked for the best recommendation method that adapted to internal data of Triptrus. This research started with gathering internal data that followed by build recommendation based on several methods. Methods that used in this research are collaborative filtering, content based filtering, hybrid filtering and stochastic gradient descent. Based on the research result, collaborative filtering, content based, and hybrid lack of capability to give good recommendation. The best result from these three methods is hybrid with an error 0.71. In the other side stochastic gradient descent could gave best recommendation with smallest error ratio RMSE at 0.11. The result of this research is recommendation model that can give best recommendation adapted to Triptrus internal data. Best recommendation model is model that generated by stochastic gradient descent."

"ABSTRAKPerkembangan jumlah pengguna internet di Indonesia mengalami peningkatan dari tahun ke tahun. Perkembangan internet berdampak pula pada munculnya beberapa ecommerce, tidak terkecuali ecommerce yang bergerak dalam jasa pemesanan tiket dan hotel. Selain itu, internet juga mendukung media sosial untuk mengekspresikan opini yang objektif tentang suatu produk/jasa. Media sosial dijadikan sebagai media electrocic word of mouth e-wom oleh pelaku jasa ecommerce. Peneltian ini terkait analisis sentiment, reputasi brand, dan jaringan sosial di Twitter terkait ecommerce yang bergerak pada bidang pemesanan hotel dan tiket. Data yang digunakan di dalam penelitian ini merupakan data yang berhubungan dengan mention @pegi_pegi, @traveloka, dan @tiket yang diambil dari periode 24 September 2016 sampai 21 November 2016. Penelitian ini menggunakan algoritme GaussianNB, MultinomialNB, BernoulliNB, ME, SVM, dan Xgboost pada proses pembuatan model. Pada kasus imbalance data, proses pembuatan model menggunakan SMOTE yang bertujuan menyeimbangkan jumlah kelas pada data yang ada. Akurasi terbaik diperoleh dengan menggunakan algoritme SVM SMOTE sebesar 0.96, presisi sebesar 0.96, recall sebesar 0.96, dan F1-Score sebesar 0.96. Nilai reputasi brand untuk @pegi_pegi sebesar -6, @traveloka sebesar -5, dan @tiket sebesar -2. Akun yang memiliki tingkat pengaruh secara keseluruhan terhadap @pegi_pegi yaitu @calvinjeremy, @traveloka yaitu @banyuwangi_kab, dan @tiket yaitu @IndahJuli.

---

ABSTRACTThe number of internet users in Indonesia has increased from year to year. Internet development impact on the emergence of e commerces, including in ticket and hotel reservation services. In addition, the internet also supports social media to express their opinions about a product service. Social media is used as a medium electrocic word of mouth e wom by actor rsquo s ecommerce services. This study focuses on sentiment analysis, brand reputation, and social networking on Twitter related to e commerce that focuses on the hotel and ticket reservations. The data used in this research is data related to pegi pegi, traveloka, and tiket taken from the period 24 September 2016 until 21 November 2016. This research uses a GaussianNB algorithm, MultinomialNB, BernoulliNB, ME, SVM, and Xgboost in the modeling process. In case of imbalanced data, process modeling using SMOTE which aims to balance the number of classes on existing data. Best accuracy obtained by using SVM algorithm SMOTE is 0.96, the precision is 0.96, the recall is 0.96, and F1 Score is 0.96. Brand reputation for pegi pegi is 6, traveloka is 5, and tiket is 2. Accounts that have effect on pegi pegi is calvinjeremy, traveloka is banyuwangi kab, and tiket is IndahJuli."

"Penerapan Teknologi Informasi dan Komunikasi (TIK) untuk meningkatkan keunggulan kompetitif.tidak hanya dimanfaatkan oleh sektor industri, namun juga sektor pemerintahan. Pemerintah Indonesia sendiri di dalam kaitannya dengan pemanfaatan TIK, telah membangun suatu sistem yang memungkinkan masyarakat untuk melaporkan keluhan dan aspirasinya melalui sistem LAPOR!. Sistem LAPOR! ciptaan pemerintah ini ternyata ditanggapi dengan antusias oleh masyarakat, terbukti dengan banyaknya laporan yang masuk ke pemerintah. Guna membantu kinerja pemerintah, dilakukan penelitian untuk menganalisis data tekstual laporan masyarakat dengan text mining untuk kemudian dilakukan disposisi otomatis ke dalam dua kategori utama LAPOR! yaitu topik dan instansi terkait. Disposisi otomatis dilakukan menggunakan teknik problem transformation pada multilabel classification melalui algoritma klasifikasi support vector machine dan naïve bayes. Hasil penelitian menunjukkan bahwa disposisi otomatis dapat diterapkan ke dalam sistem LAPOR! dan dapat meningkatkan kinerja disposisi laporan. Algoritma yang menghasilkan performa terbaik di dalam penerapannya adalah algoritma support vector machine

---

The application of Information Technology and Communication (ICT) to escalate the competitive advantage is not only used in the industrial sector, but also in the government as well. The government of the Republic of Indonesia itsef, in the use of ICT, has built a system that enable its citizen to report their grievance and aspiration through LAPOR! system. This system turned out to be accepted with great enthusiasm by the public, as evidenced by the many reports to the government. In order to support the government’s performance, research is conducted to analyze the textual data using text mining, for later automatic disposition into two groups of LAPOR!'s category which is topik and instansi terkait. disposition is done using problem transformation technique in multilabel classification through support vector machine and naïve bayes classification algorithm. The result showed that automatic disposition can be applied into LAPOR! system and improves the report disposition’s performance. Algorithm that produces the best performance in the application is support vector machine. "

"Karya ini menggunakan encoder visual berbasis Concept-Map yang menanggulangi masalah penghubungan informasi dari citra yang telah di-encode oleh sebuah jaringan saraf tiruan konvolusional ke dalam ranah semantik yang diproses oleh jarinagn saraf tiruan berbasis waktu. Pendekatan ini menggunakan komponen attention visual yang mengembangkan jaringan konvolusional sebelum dipropagasi ke jaringan berbasis waktu. Untuk meningkatkan pembelajaran cross-entropy, model dilatih dengan metode reinforcement learning dengan cara melatih value dan policy network berdasarkan jarak visual-semantic embedding distance dari representasi vector sebagai sinyal reward. Visual-semantic embedding space belajar dan menghasilkan vector space untuk citra dan teks, lalu digunakan sebagai tolak ukur qualitas suatu teks yang mendeskripsikan suatu citra. Sinyal reward membantu mengarahkan dan memaksimalkan probabilitas suatu deskripsi bagus muncul. Dataset yang digunakan adalah Flickr8k dan metric yang dilaporkan adalah BLEU-1 hingga BLEU-4

---

This work features a Concept-Map visual encoder that tackles the issue of linking encoded image information from convolutional neural networks to semantic domain processed by recurrent neural networks. The approach utilizes visual attention that extends the convolutional network before being propagated to through the recurrent network. To improve upon cross entropy learning, the model is then trained on reinforcement learning by training a value and policy network on visual-semantic embedding distance of vector representations as reward signals. The visual-semantic embedding space that jointly learns a common vector space for encoding image and caption is used to measure the quality of generated caption computing how close the vector representation is to the vector representation of the input image. The reward signal guides the policy to maximize the probability of producing good captions. The Flickr8K dataset is used and BLEU-1 to BLEU-4 for is reported."

"Alzheimer dikategorikan sebagai salah satu dimensia berat dengan bentuk otak yang mengalami penyusutan dan volume otak berkurang secara keseluruhan. Selain itu, Alzheimer juga mengakibatkan terjadinya atrophy pada bagian hippocampus. Korelasi antara penyusutan bentuk otak dan berkurangnya volume juga mempengaruhi perubahan bentuk tekstur. Pada penelitian yang diusulkan, perluasan dari Local Binary Pattern (LBP) sebagai ekstraksi fitur diperkenalkan. Complete Local Binary Pattern of Sign and Magnitude (CLBPSM) dan Complete Local Binary Pattern of Sign and Magnitude from Three Orthogonal Planes (CLBPSM-TOP) diperkenalkan sebagai deskriptor ekstraksi fitur 2D dan 3D. Dikarenakan fitur yang begitu banyak dihasilkan, maka Principal Component Analysis (PCA), kernel PCA dan Factor Analysis (FA) digunakan sebagai salah satu metode seleksi fitur. Selanjutnya, lima buah classifier digunakan untuk klasifikasi binary class dan multiclass dari Alzheimer, mild cognitive impairment dan normal pada bagian keseluruhan otak dan hippocampus. Hasil eksperimen dengan tiga buah skenario menunjukkan bahwa metode CLBPSM dan CLBPSMTOP mampu memberikan hasil akurasi dan performance yang lain dengan nilai rata-rata antara 70% - 100% untuk bagian keseluruhan otak dan hippocampus. Pendekatan CLBPSM-TOP sebagai deskriptor 3D juga menggungguli metode LBP-TOP pada studi literatur dengan rata-rata kenaikan akurasi sebesar 30% untuk semua klasifikasi.

---

Alzheimer`s disease is categorized as one of heavy dementia with the brain forms will shrink, and reduced overall volume of brain. The correlation between brain form shrinkage and reduction of brain volume also affect deformation texture. In the proposed research, the expansion of local binary pattern (LBP) as feature extraction method is introduced. Complete local binary pattern of sign and magnitude (CLBPSM) and complete local binary pattern of sign and magnitude from three orthogonal planes (CLBPSM-TOP) are introduced as a 2D and 3D feature extraction descriptor. Due to so many features are generated, then the principal component analysis (PCA), kernel PCA and Factor Analysis (FA) are used as a method of feature selection. Furthermore, five classifiers are used for binary class and multiclass classification of Alzheimer's, mild cognitive impairment and normal in the whole brain and hippocampus. The experimental results with three scenarios show that CLBPSM and CLBPSM-TOP methods are able to provide accuracy and the other performance results with an average value between of 70% - 100% for the whole brain and hippocampus. CLBPSM-TOP approach as a 3D descriptor also outperformed LBP-TOP method from the previous study with an average accuracy increase 30% for all classifications."

"[ABSTRAKRontgen dada atau Chest X-Ray (CXR) merupakan salah satu aplikasi pencitraan medis yang paling sering digunakan dalam pendeteksian kelainan khususnya tumor pada paru – paru. Untuk menentukan diagnosis kelainan tersebut, seorang dokter masih mengandalkan pengamatan visual dalam pembacaan hasil citra CXR sehingga penilaian bersifat subyektif tergantung pada masing – masing dokter. Oleh karena itu, pada penelitian ini dilakukan perancangan sistem pengolahan citra sebagai alat bantu identifikasi kelainan paru – paru. Kategori citra CXR yang digunakan adalah citra pada keadaan normal, tumor, dan kelainan bukan tumor. Tahapan pengolahan yang dilakukan berupa pre-processing menggunakan median filtering dan ekualisasi histogram serta proses segmentasi menggunakan otsu’s thresholding dan active contour : snake. Uji hasil pengolahan citra dengan hasil diagnosis dokter menggunakan jaringan syaraf tiruan backpropagation menghasilkan akurasi sebesar 92,85 %.

---

ABSTRACTChest X-Ray (CXR) is a medical imaging applications that most commonly used for detects of abnormalities, especially tumors of the lung. To determine the abnormality diagnosis, doctors still rely on visual observations to read a CXR image, so that the assessments are subjective depending on each doctor. This study purposes to design an image processing system as a tool for identification of lung’s abnormalities. It used three classification of CXR image, which are lungs image in normal circumstances, tumors, and abnormalities besides tumor. Stages of image processing are done in the form of pre-processing using a median filtering and histogram equalization and also the process of segmentation using Otsu's thresholding and active contour: snake. Test the image processing results with the results of the doctor's diagnosis using artificial neural network backpropagation produces an accuracy of 92,85 %., Chest X-Ray (CXR) is a medical imaging applications that most commonly used for detects of abnormalities, especially tumors of the lung. To determine the abnormality diagnosis, doctors still rely on visual observations to read a CXR image, so that the assessments are subjective depending on each doctor. This study purposes to design an image processing system as a tool for identification of lung’s abnormalities. It used three classification of CXR image, which are lungs image in normal circumstances, tumors, and abnormalities besides tumor. Stages of image processing are done in the form of pre-processing using a median filtering and histogram equalization and also the process of segmentation using Otsu's thresholding and active contour: snake. Test the image processing results with the results of the doctor's diagnosis using artificial neural network backpropagation produces an accuracy of 92,85 %.]"