Articles
Optimasi Komposisi Pakan Sapi Perah Menggunakan Algoritma Genetika
Durrotul Fakhiroh;
Wayan Firdaus Mahmudy;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 1 No 1 (2017): Januari 2017
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (802.777 KB)
Hambatan terbesar yang dialami oleh peternak sapi perah adalah penggunaan komposisi pakan yang tidak efisien. Dalam sudut pandang ekonomi, biaya untuk pembelian pakan ternak merupakan biaya tertinggi dalam usaha peternakan, sehingga harus ditekan serendah mungkin untuk memaksimalkan pendapatan dengan tetap memperhatikan nutrisi yang dibutuhkan oleh sapi perah. Agar dapat mencapai dua hal tersebut dilakukan optimasi terhadap ransum agar dapat memenuhi kebutuhan nutrisi dengan biaya yang minimal. Algoritma genetika merupakan salah satu metode yang sesuai untuk memecahkan permasalahan optimasi. Representasi yang digunakan adalah real code dimana setiap kromosom mewakili bobot dari bahan pakan, dan panjang kromosom tergantung dari banyaknya bahan pakan. Metode crossover yang digunakan adalah extended intermediete, proses mutasi menggunakan metode random mutation, sedangkan elitism adalah metode yang digunakan dalam proses seleksi. Berdasarkan hasil pengujian yang telah dilakukan, diperoleh parameter optimal yaitu pada populasi 100, generasi 200, serta kombinasi cr dan mr sebesar 0.3 dan 0.3. Hasil akhir yang didapatkan berupa rekomendasi komposisi ransum dengan biaya yang minimal dan kebutuhan nutrisi sapi perah tetap terpenuhi.
Deteksi Autisme pada Anak Menggunakan Metode Modified K-Nearest Neighbor (MKNN)
Zahra Swastika Putri;
Rekyan Regasari Mardi Putri;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 1 No 3 (2017): Maret 2017
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1123.03 KB)
Autism is a childhood and developmental disorder that characterized by lack of communication, cognition, imagination and social interaction activities. Many people didn't recognize the symptoms of autism disorder until the first three or seven years of life. Delay, similarities of symptoms and lack of knowledge about autism cause imprecision treatment handling, and increased number of sufferers. Identification of autism differentiated into severe autism, moderate autism, mild autism and non- autism. Modified K-Nearest Neighbor (MKNN) method is a method that enhancing performance of conventional K-Nearest Neighbor method. There're validity of the train data process and weight voting process to robust neighbors of training dataset and strengthen the performance results. Based on variant value of k testing obtained 83.33% accuracy at dissimilarity measure. Based on composition of balance training data testing obtained 90% accuracy at euclidean distance. Based on amount of training data testing obtained 79.17% average accuracy. Based on variation of training data testing obtained 83.33% accuracy at dissimilarity measure. Based on results of such testing accuracy, pointed out that the detection of children's autism using MKNN method have a pretty good degree of accuracy and capable to classify and detection the autism symptoms based on perceived symptoms user input.
Penerapan Metode K-Nearest Neighbor (KNN) dan Metode Weighted Product (WP) Dalam Penerimaan Calon Guru Dan Karyawan Tata Usaha Baru Berwawasan Teknologi (Studi Kasus : Sekolah Menengah Kejuruan Muhammadiyah 2 Kediri)
Nihru Nafi' Dzikrulloh;
Indriati Indriati;
Budi Darma Setiawan
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 1 No 5 (2017): Mei 2017
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (878.441 KB)
World of particular employment agencies Vocational High School, many a teacher or school employee who less clever in technology of the current technological developments. Actually, it is in need of teachers and school administration employees who have qualified human resources high in the knowledge of science and technology. The school is in need it is because it affects how do learning on students in school. To meet the desired standards of quality teachers, during The Vocational High School Muhammadiyah 2 Kediri is selection and recruitment of teachers by means of manual employees. The selection has been done manually through the test phase 4 aspects of your application letter and attachments GPA averages, academic test, test general knowledge of science and technology (IPTEK), and interview. The data collection process for the selection still use manual. Therefore, we need a web-based system so that the selection acceptance of new teacher candidates can run more effectively and efficiently. On this website using K-Nearest Neighbor (KNN) and the method of Weighted Product (WP). K-Nearest Neighbor used to determine the weight of each criterion to classify the good or bad. After classifying the KNN method, the selection of prospective teachers will be recruited by the school Vocational High School Muhammadiyah 2 Kediri using Weight Product (WP). Weight Product used to determine the results of the classification by KNN method to perform a ranking in order to take the best results. Tests conducted consisting of, testing the accuracy of the value of K means and accuracy testing of the WP value criteria weighting method. The accuracy of the test results obtained suitability accuracy value by 94%, precision 80%, and recall 80%.
Klasifikasi Tweets Pada Twitter Dengan Menggunakan Metode Fuzzy K-Nearest Neighbour (Fuzzy K-NN) dan Query Expansion Berbasis Apriori
Joda Pahlawan Romadhona Tanjung;
Mochammad Ali Fauzi;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 1 No 5 (2017): Mei 2017
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1372.995 KB)
Twitter is a unique conversation tool that allows us to send and receive short messages called tweets in the Twitter community. Tweets are short messages that have a length of 140 characters. Tweets that appear on the homepage are all jumbled into one, posted variety ranging from the economy, sports, technology, automotive, healthcare and others. When users search for a news or information desired, the problem that arises is Twitter user difficult to find tweets. The classification process can be performed to categorize a tweets using an algorithm Fuzzy K-Nearest Neighbour. However, the process of classifying a tweets it is difficult to do because the tweets in the form of short-text. Therefore, before doing the classification process a tweets done preprocessing and word expansion beforehand with Query Expansion algorithms in order to provide maximum results in the classification. In the study conducted to produce the best accuracy by 82%. Best accuracy is obtained when using the Fuzzy KNN method with Query Expansion without preprocessing and threshold for the support value> = 0.15 and the value of confidence> = 1.
Pembangkitan Nilai Belief Pada Dempster-Shafer Dengan Particle Swarm Optimization (PSO) Untuk Penentuan Pasal Kasus Penganiayaan
Merry Gricelya Nababan;
Rekyan Regasari Mardi Putri;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 1 No 10 (2017): Oktober 2017
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1004.216 KB)
The crime against the body and life continues to increase every year, judges as decision makers against criminal defendants have a very important role in providing decisions. However, there are some things that the judge needs to consider in making decisions, so that the problem of uncertainty can be a judge's obstacle. The author applies a method that can solve the problem of this uncertainty is Dempster-shafer (D-S). D-S algorithm has belief value that serves to determine the influence between symptoms obtained from an expert. In this case the expert can not give the value of belief karana must be in accordance with the evidence and real sanctions. So with Particle Swarm Optimization algorithm (PSO) belief value will be raised as well as doing optimization to get maximum results. In accordance with the test conducted from the case data of the penganiaayan obtained maximum belief value based on PSO parameter test. The result of system accuracy calculation by using belief value that has been optimized with D-S on 29 cases of abuse shows accuracy of 13.79%. The result of this accuracy is not maximal due to complex problems with the output (Output) of the system more than one. For further research, we can use Artificial Neural Network (ANN) method or with algorithm Analytic Hierarchy Process (AHP).
Prediksi Kebutuhan Air PDAM Kota Malang Menggunakan Metode Fuzzy Time Series Dengan Algoritma Genetika
Khaira Istiqara;
Muhammad Tanzil Furqon;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 1 (2018): Januari 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1140.093 KB)
Water is one of the basic needs of living things derived from natural resources. The Government provides a regional water company called Perusahaan Daerah Air Minum (PDAM) to fullfil the clean water needs of the people of Indonesia, one of which is located in Malang. PDAM water needs prediction system serves to predict the water needs of the people of Malang, so water needs will be guaranteed in the future. Variable used is PDAM water usage data from 2008-2013. Genetic algorithms are used to optimize the subset of universe in fuzzy time series. Search solution uses real-coded chromosome representation, then processed with genetic operator (crossover, mutation and selection). Method of genetic operator used is one-cut-point crossover, uniform mutation and elitism selection. The result of testing genetic algorithm parameter values, obtained the optimal population size is 360, the length of chromosome is 60, the best combination of crossover rate and mutation rate are 0.4 and 0.2, and the number of optimal generation is 550. Based on the best genetic algorithm parameter value, obtained the prediction result with the error value (MAPE) is 2.266776%. These results showed a good predictive ability with low error values.
Analisis Sentimen Dengan Query Expansion Pada Review Aplikasi M-Banking Menggunakan Metode Fuzzy K-Nearest Neighbor (Fuzzy k-NN)
Nanda Cahyo Wirawan;
Indriati Indriati;
Putra Pandu Adikara
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 1 (2018): Januari 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1099.025 KB)
In this digital era, bussiness grow significantly by using digital application. Banking is one field of business that utilizes the current technological advances very well. Mobile banking is one of the most popular digital banking products, because it is not as complicated as SMS banking or internet banking. In order to face the strict banking business, every company applying feedback from their customers. Now customers can use the review feature that provided by apps store. There's a lot of reviews that received every day, and it takes some time to knowing what kind of review is that. Systems with machine learning are expected to save time to sort out textual data that containing polarity. The system's machine learning in this study was made using fuzzy k-nearest neighbor (fuzzy k-NN) method. The fuzzy k-NN method is a combined method between fuzzy logic and the k-Nearest Neighbor algorithm. The weighting method for processing textual data into numerical data that can be computed is using TF-IDF method with Cosine similarity to calculate the distance between data. The output of this system is the classified data review. Based on the results of the tests, this system produces the best F-Measure is 0.9273 and the worst is 0.8349.
Klasifikasi Dokumen Tumbuhan Obat Menggunakan Metode Improved K-Nearest Neighbor
Arinda Ayu Puspitasari;
Edy Santoso;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 2 (2018): Februari 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1105.725 KB)
The high utilization rates of medicinal plants is leading to increase the studies on it. Those studies certainly require documentation that contains information about medicinal plants. The large and scattered documentation cause difficulties in searching for information about medicinal plants. To overcome these problems a system that can classify the document automatically is needed to make the information search work more effective and efficient. K-Nearest Neighbor is the algorithm often used to classify text, but has a weakness in accuracy because of the fixed k values for each category. K values is the amount of the closest training data to the test data. Improved k-Nearest Neighbour is the algorithm used in this study to overcome the problem where the different k values will be applied based on the amount of the training data for each category. The average accuracy for the k values testing is 70,99%. The training data variation testing shows that the bigger amount of training data the higher average accuracy will be. The unbalanced data testing showed that the balance data training category has 1,9% better accuracy than the unbalanced category.
Implementasi Metode Analytic Hierarchy Process - Weighted Product Untuk Rekomendasi Hunian Ideal (Studi Kasus: Kota Malang)
Rizaldy Aditya Nugraha;
Indriati Indriati;
Imam Cholissodin
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 2 (2018): Februari 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1666.731 KB)
The purpose of this study is to help prospective house buyers in getting the recommendation for an ideal house to be purchased. Prospective house buyers that were looking for a house of their dreams still found it difficult to obtain the appropriate recommendation suitable with their desires. Therefore, this study was conducted to create a decision support system application of ideal house recommendation to facilitate a prospective house buyer in obtaining an ideal house recommendation. The input data used on this system is a weight priority measure for each criteria and sub criteria of the house specified by the prospective house buyer. Then these input data are calculated by using analytic hierarchy process - weighted product method. The analytic hierarchy method is used to obtain the criteria and sub criteria weight which is then used for the calculation of weighted product method. The final result of this system is the rank order of ideal house recommendation. The test performed on this system is done on the pairwise comparison matrices with 80% accuracy.
Penerapan Sentimen Analisis Acara Televisi Pada Twitter Menggunakan Support Vector Machine dan Algoritma Genetika sebagai Metode Seleksi Fitur
I Made Budi Surya Darma;
Rizal Setya Perdana;
Indriati Indriati
Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer Vol 2 No 3 (2018): Maret 2018
Publisher : Fakultas Ilmu Komputer (FILKOM), Universitas Brawijaya
Show Abstract
|
Download Original
|
Original Source
|
Check in Google Scholar
|
Full PDF (1037.703 KB)
Rating is one approach method that can be used to find out about audience satisfaction of a TV show. In Indonesia, rating was calculated by using AGB Nielsen services. However, rating that AGB Nielsen produced was based on the people watching bahavior in 10 major cities in Indonesia. Therefore, rating in Indonesia requires another method to get the watching behavior of the whole people in Indonesia. Twitter, can be used to get Indonesia people watching behavior. Through the published tweets, it can be applied the process of extracting information by using classification techniques to get the opinions. One of the classification techniques that can be applied to text categorization is the Support Vector Machine (SVM) it`s suitable for multiple dimension data. By optimizing the features that will be used, it can provide optimal results with less features used. One of the feature selection methods that can be applied to SVM is the genetic algorithm (GA). System calculates the rating, based on positive and negative sentiments about the TV show and divided by the population of the tweet used. The rating comparison test that produced by AGB Nielsen and system shows an average error value of 0.562. In testing the accuracy before and after the feature selection method is applied, showed results with average error value 0.62%.