(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Statistical method for detecting trend and seasonality in non stationary time series data

Statistical method for detecting trend and seasonality in non stationary time series data / Muhammad Danial Zaharudin
Penyelidikan ini dilakukan untuk mencadangkan teknik statistik yang boleh digunakan untuk mengenal pasti perubahan aliran dan bermusim didalam siri masa. Ia juga dikaji untuk menambah baik penyelidikan yang dilakukan oleh Muhammad Farid [7]. Muhammad Farid [7] telah menghasilkan kerangka kerja dan perisian untuk analisis dan peramalan siri masa yang mempunyai aliran dan bermusim. Walau bagaimana pun, kerangka kerja yang telah dibuat oleh beliau tidak termasuk keadah untuk mengesan aliran dan bermusim. Didalam kajian ini, dua kaedah yang dicadangkan untuk mengenal pasti perubahan aliran adalah simple moving average dan regression analysis. Terdapat lima kaedah yang dicadangkan untuk mengesan bermusim iaitu average graph, autocorrelation plot, correlation analysis, lag plot dan detrended graph. Kaedah yang terakhir telah digunakan untuk memberi kata putus untuk kes ketidak samaan dalam keputusan yang diperolehi dari empat kaedah yang dicadangkan. Walaupun kelima-lima kaedah ada di nyatakan didalam kesusasteraan tetapi average graph dan correlation analysis telah dicadangkan untuk digunakan pertama kali nya untuk mengesan bermusim didalam kajian ini. Tiga belas kumpulan data dari pelbagai bidang telah digunakan untuk menguji kaedah-kaedah yang dicadangkan. Antara kumpulan data ini adalah dari bidang pertanian, penjana kuasa, kewangan dan jualan Semua kaedah telah menggunakan Excel-base template untuk melaksanakan kaedah tersebut kecuali kaedah autocorrelation plot. Untuk menghasilkan autocorrelation plot, perisian Minitab telah digunakan.Kaedah-kaedah ini menghasilkan graf dan rumusan statistik seperti autocorrelation coefficient dan correlation coefficent. Kajian ini turut menunjukkan cara untuk menafsir graf dan pekali. Di peringkat terakhir kajian ini, kesemua kaedah ini telah dimasukkan kedalam kerangka kerja Muhammad Farid [7]. _______________________________________________________________________________________________________ This research investigated statistical methods that can be used to detect trend and seasonality in time series data. It also improved the work that had been done by Muhammad Farid [7]. Muhammad Farid [7] had developed a framework and software for forecasting time series data that have trend and seasonality. However, the framework did not include methods for detecting trend and seasonality. In this research, two methods had been suggested for trend; simple moving average and regression analysis. Five methods had been suggested for seasonality; average graph, autocorrelation plot, correlation analysis, lag plot, and detrended graph. The latter was used to confirm seasonality in the case of conflicting results from the first four methods. Though all of the suggested methods are documented in the literature, the use of average graph and correlation analysis for seasonality are suggested for the first time in this research. Thirteen time series data sets from various fields such as agriculture, power utility, finance and sales were used to evaluate the methods. Excel-based templates for executing all of the methods except for autocorrelation plot were developed. Minitab software was used to generate the autocorrelation plot. The methods produced graphs and statistical summaries such as autocorrelation coefficients and correlation coefficients. This research also shows how to interpret the graphs and the coefficients. In the final part of this research, the methods were integrated into Muhamad Farid’s framework.
Contributor(s):
Muhammad Danial Zaharudin - Author
Primary Item Type:
Final Year Project
Identifiers:
Accession Number : 875004638
Barcode : 00003093641
Language:
English
Subject Keywords:
statistical methods; framework; autocorrelation plot
First presented to the public:
1/1/2012
Original Publication Date:
3/20/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 131
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-03-20 13:07:40.66
Date Last Updated
2019-01-07 11:24:32.9118
Submitter:
Mohd Jasnizam Mohd Salleh

All Versions

Thumbnail Name Version Created Date
Statistical method for detecting trend and seasonality in non stationary time series data1 2018-03-20 13:07:40.66