|
The Imputation Many Missing Value in Time Series Data Use Multivariate Relationships |
|---|---|
| รหัสดีโอไอ | |
| Creator | พยุง มีสัจ |
| Title | The Imputation Many Missing Value in Time Series Data Use Multivariate Relationships |
| Contributor | กรสิริณัฐ โรจนวรรณ์ |
| Publisher | Faculty of Information Science and Technology, Mahanakorn University of Technology |
| Publication Year | 2561 |
| Journal Title | Journal of Information Science and Technology |
| Journal Vol. | 8 |
| Journal No. | 1 |
| Page no. | 16-25 |
| Keyword | Imputation value, Time series, Multivariate data, Artificial neural network, Data mining, Learn Machine |
| URL Website | https://tci-thaijo.org/index.php/JIST |
| Website title | Journal of Information Science and Technology |
| ISSN | 2651-1053 |
| Abstract | Time series information is important in many tasks. This is useful for predicting trends in business judgment. The problem of data collection is losing many potential data in the field. This make data analyzing cannot be performed efficiently. This paper proposes how to fill in the missing time value in the time series data. Many use multi -variable relationships by using existing data, create a fill missing value. The research focused on finding the right data model for teaching the missing value model by comparison of four alternative techniques: Row Average, K-Nearest Neighbor(KNN), Fuzzy Logic Systems and Artificial Neural Network .The research found Artificial Neural Network technique provides predictive effect a nd when used to replace lost values, the results are similar. T hat using some of the available data from multiple variables. It can be used to create a lossless representation. Variable data many variables will directly affect the formatting of data and models. Good data before creating a replacement for lost values. |
