ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION
The era of big data encourages the application of streaming data in its analysis and modeling. In this thesis, a model for streaming space-time data is developed using the Generalized STAR or GSTAR model as backbone, hereinafter referred to as Online GSTAR. Using the regret minimization technique, t...
Saved in:
| Main Author: | |
|---|---|
| Format: | Theses |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/65016 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| id |
id-itb.:65016 |
|---|---|
| spelling |
id-itb.:650162022-06-20T09:06:15ZONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION Uzila Dwiyanda, Albers Indonesia Theses generalized space-time autoregressive, online learning, regret, aggregating algorithm, streaming data INSTITUT TEKNOLOGI BANDUNG https://digilib.itb.ac.id/gdl/view/65016 The era of big data encourages the application of streaming data in its analysis and modeling. In this thesis, a model for streaming space-time data is developed using the Generalized STAR or GSTAR model as backbone, hereinafter referred to as Online GSTAR. Using the regret minimization technique, the two online models GSTAR-OGD and GSTAR-ONS were compared with the GSTAR model. It was found that the Online GSTAR is more robust to changes in the nature of the data over time than the GSTAR model. In addition, unlike GSTAR, the Online GSTAR model has at most constant training and prediction time and it doesn’t depend on the amount of data that’s been read. Lastly, the performance of Online GSTAR model is asymptotically approaching the best model’s performance. Further development of Online GSTAR with Aggregating Algorithm is able to eliminate the need for several initial observations to determine the order of the model. text |
| institution |
Institut Teknologi Bandung |
| building |
Institut Teknologi Bandung Library |
| continent |
Asia |
| country |
Indonesia Indonesia |
| content_provider |
Institut Teknologi Bandung |
| collection |
Digital ITB |
| language |
Indonesia |
| description |
The era of big data encourages the application of streaming data in its analysis and modeling. In this thesis, a model for streaming space-time data is developed using the Generalized STAR or GSTAR model as backbone, hereinafter referred to as Online GSTAR. Using the regret minimization technique, the two online models GSTAR-OGD and GSTAR-ONS were compared with the GSTAR model. It was found that the Online GSTAR is more robust to changes in the nature of the data over time than the GSTAR model. In addition, unlike GSTAR, the Online GSTAR model has at most constant training and prediction time and it doesn’t depend on the amount of data that’s been read. Lastly, the performance of Online GSTAR model is asymptotically approaching the best model’s performance. Further development of Online GSTAR with Aggregating Algorithm is able to eliminate the need for several initial observations to determine the order of the model. |
| format |
Theses |
| author |
Uzila Dwiyanda, Albers |
| spellingShingle |
Uzila Dwiyanda, Albers ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION |
| author_facet |
Uzila Dwiyanda, Albers |
| author_sort |
Uzila Dwiyanda, Albers |
| title |
ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION |
| title_short |
ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION |
| title_full |
ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION |
| title_fullStr |
ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION |
| title_full_unstemmed |
ONLINE GSTAR ALGORITHM FOR STREAMING SPACE-TIME DATA PREDICTION |
| title_sort |
online gstar algorithm for streaming space-time data prediction |
| url |
https://digilib.itb.ac.id/gdl/view/65016 |
| _version_ |
1822004732294594560 |