Peramalan Data Intermiten Curah Hujan Menggunakan Metode Penggabungan Arima Dan Group Method Of Data Handling (GMDH) (Studi Kasus: Stasiun Meteorologi Perak)
Bibliographic record
Abstract
Penelitian sebelumnya terkait peramalan menggunakan metode penggabungan ARIMA dan GMDH yang dilakukan oleh Shabri dan Samsudin telah menghasilkan peramalan yang efektif pada 114 data observasi Canadian Lynx. Hasil Error yang dihasilkan cenderung kecil dibandingkan menggunakan metode ARIMA ataupun Artifical Neural Network (ANN.) Penelitian kali ini, ingin mencoba menerapkan metode yang sama pada intermittent data pada data timeseries curah hujan di daerah Surabaya. Surabaya merupakan kota yang mempunyai curah hujan rata-rata 300-400mm. Banyak daerah di Surabaya yang terkena dampak banjir dikarenakan tingginya curah hujan tersebut. Untuk menanggulangi hal tersebut, perlu adanya prediksi yang tepat akan terjadinya curah hujan yang tidak konsisten. Data yang tidak konsisten atau banyak nilai 0 (nol)-nya dapat dikatakan sebagai intermittent data. Sebelum melakukan penggabungan metode ARIMA dan GMDH, dilakukan pengelolahan data dengan menggunakan metode Bootsrap. Bootstrap merupakan salah satu metode dalam melakukan transformasi data untuk data intermiten. Metode ini mampu menghasilkan error yang kecil dibandingkan dengan metode-metode lain. beberapa model yaitu model ARIMA,ARIMA-ARCH, GMDH, dan Hybrid ARIMA-GMDH. Dengan memperhitungkan Symmetric Absolute Mean Percentage Error (SMAPE) pada masing-masing model didapatkan model terbaik yakni model Hybrid ARIMA-GMDH. Dengan data input hasil residu peramalan ARIMA dengan data bootstrap dihasilkan SMAPE data training sebesar 0,0322% dan data testing sebesar 0.0138%. Dibandingkan dengan SMAPE data testing model lain yakni ARIMA sebesar 0.0287%, ARCH sebesar 0.0391%, dan GMDH sebesar 0.0429%. Hasil peramalan ini diharapkan dapat membantu pihak terkait dalam mempersiapkan langkah antisipasi dari kejadian yang dapat ditimbulkan saat curah hujan tinggi.=================== Recently studies about forecasting using hybrid ARIMA-GMDH method has performed by Shabri and Samsudin resulting an effective forecast using 114 Canadian Lynx data as observational data. The result of Error within hybrid ARIMA-GMDH generated tends to be smaller than individual method like ARIMA or Artifical Neural Network (ANN) method. This research is trying to apply the same method to intermittent time series data like rainfall in Surabaya areas. City of Surabaya has 300-400mm average of rainfall. Many areas in Surabaya are affected by this high rainfall caused floods. A solutions of this problems, we should have a precise predictions of that incosistent rainfall. Incosistent data or many zero values can be refered as intermittent data. Before we perform hybrid ARIMA-GMDH method, the data should be managed by Bootstraping Method. Bootstrap is the one methods of transforming data for intermittent data. This method has capability to generating small Errors compared to other methods. This research tested several models of ARIMA, ARIMA-ARCH, GMDH, and Hybrid ARIMA-GMDH models. Using Symmetric Absolute Mean Percentage Error (SMAPE) the error of each models has calculated, Hybrid ARIMA-GMDH have selected as best model. With input data of forecasting residu of ARIMA with bootstrap data resulting error of data training 0.0322% and data testing 0.0138%. This error compared with SMAPE other testing data of each models like ARIMA is 0.0287%, ARCH is 0.0391%, and GMDH is 0.429%. The forecasting result of this research maybe can help another concerned parties in anticipating some emergency events caused by the high rainfall.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.003 | 0.006 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.003 |
| Science and technology studies | 0.001 | 0.001 |
| Scholarly communication | 0.003 | 0.002 |
| Open science | 0.002 | 0.002 |
| Research integrity | 0.001 | 0.002 |
| Insufficient payload (model declined to judge) | 0.007 | 0.003 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".