Claim Missing Document
Check
Articles

Found 3 Documents
Search

Regression Modelling for Precipitation Prediction Using Genetic Algorithms Asyrofa Rahmi; Wayan Firdaus Mahmudy
TELKOMNIKA (Telecommunication Computing Electronics and Control) Vol 15, No 3: September 2017
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.12928/telkomnika.v15i3.4028

Abstract

This paper discusses the formation of an appropriate regression model in precipitation prediction. Precipitation prediction has a major influence to multiply the agricultural production of potatoes in Tengger, East Java, Indonesia. Periodically, the precipitation has non-linear patterns. By using a non-linear approach, the prediction of precipitation produces more accurate results. Genetic algorithm (GA) functioning chooses precipitation period which forms the best model. To prevent early convergence, testing the best combination value of crossover rate and mutation rate is done. To test the accuracy of the predicted results are used Root Mean Square Error (RMSE) as a benchmark. Based on the RMSE value of each method on every location, prediction using GA-Non-Linear Regression is better than Fuzzy Tsukamoto for each location. Compared to Generalized Space-Time Autoregressive-Seemingly Unrelated Regression (GSTAR-SUR), precipitation prediction using GA is better. This has been proved that for 3 locations GA is superior and on 1 location, GA has the least value of deviation level.
Profit Optimization Based on Total Production in Textile Home Industry Using Evolution Strategies Algorithms Mabafasa Al Khuluqi; Wayan Firdaus Mahmudy; Asyrofa Rahmi
e-2477-1929
Publisher : Institute of Research and Community Service, University of Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (900.795 KB) | DOI: 10.21776/ub.ijleg.2016.002.02.2

Abstract

 Profit optimization became one of the main goals of the production process of home industry. A maximum profit can be achieved with proper planning of production. In the process of implementation, production planning has many constraints such as lot sizing, limited stock, overtime work and the many products derived from the same source. To address the problem, we develop computer software using a heuristic method called evolution strategy (ES). ES has capability to solve optimization problems with nearly optimal results. The result of the calculation process of evolution strategy algorithm was compared with data from the source. Computational analysis shows that ES produces a production plan that has profit of Rp 5,324,000. It is bigger than manual production plan that has profit of Rp 2,747,000. Keywords: home industry, profit, optimization, production, evolution strategies 
Hibridisasi Algoritma Genetika Dengan Variable Neighborhood Search (VNS) Pada Optimasi Biaya Distribusi Asyrofa Rahmi; Wayan Firdaus Mahmudy; Syaiful Anam
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 4 No 2: Juni 2017
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | Full PDF (987.611 KB) | DOI: 10.25126/jtiik.201742287

Abstract

AbstrakProses distribusi dianggap sangat penting bagi perusahaan karena menjadi salah satu faktor yang mempengaruhi perolehan keuntungan. Besarnya biaya yang dikeluarkan serta kompleksnya permasalahan dalam proses distribusi menjadikan permasalahan distribusi sebagai topik yang perlu diteliti lebih mendalam lagi. Karena algoritma genetika (AG) sudah terbukti mampu memberikan solusi terbaik pada berbagai macam permasalahan optimasi dan kombinatorial, maka algoritma ini digunakan untuk menyelesaikan permasalahan distribusi pada penelitian ini. Namun, penerapan GA klasik memiliki kekurangan yaitu belum mencapai titik optimum global sehingga perlu dihibridisasi menggunakan algoritma variable neighborhood search (VNS). Algoritma ini dipilih karena selain mencari solusi secara global, algoritma ini juga mencari solusi secara lokal sehingga mampu menutupi kekurangan dari GA. Dengan menggunakan hibridisasi GA dengan VNS maka biaya yang diperoleh adalah 32392960 yang dibuktikan dengan penghematan biaya sebesar 323190 jika dibandingkan dengan GA klasik yaitu 32716150. Namun, dilihat dari waktu komputasi, GA-VNS membutuhkan waktu yang relatif sama dengan GA klasik yaitu 279332 ms (milisecond) dan 265091 ms.Kata kunci: distribusi, algoritma genetika, variable neighborhood searchAbstractThe distribution process is considered importantly for the company as one of the factors that affects profitability. The costs incurred as well as the complexity of the distribution problems makes the distribution problems as a topic that need to be examined more deeply. Since the wide range of combinatorial and optimization problems have been ever solved by using genetic algorithm (GA) well then it is used to resolve the distribution problems in this study. However, the implementation of classical GA has the disadvantage that has not yet reached the global optimum so that needs to be hybridized by using variable neighborhood search (VNS) algorithm. The VNS algorithm has been chosen because its ability either to search the global solutions or local solutions. The local search of VNS algorithm is able to cover the shortage of the GA. By using hibridization of GA with VNS, the cost accrued is 32392960 as evidenced by cost savings of 323190 in comparison with the classical GA is 32716150. However, the computational time of GA-VNS is equal to its classical GA relatively.Keywords: distribution, genetic algorithm, variable neighborhood search