Some extensions to the use of randomized quantile residuals in generalized linear models
碩士 === 國立中正大學 === 數理統計研究所 === 88 === Residuals play an important role in model adequacy checking. However, graphical diagnostics based on conventional residuals may have problems for sparse data as were pointed out by Dunn and Smyth (1996). They propose...
| Main Author: | |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Published: |
2000
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/20142069927919244068 |
| id |
ndltd-TW-088CCU00477005 |
|---|---|
| record_format |
oai_dc |
| spelling |
ndltd-TW-088CCU004770052015-10-13T11:50:50Z http://ndltd.ncl.edu.tw/handle/20142069927919244068 Some extensions to the use of randomized quantile residuals in generalized linear models 隨機分位數殘差值用於廣義線性模型之一些推廣 郭籠慶 碩士 國立中正大學 數理統計研究所 88 Residuals play an important role in model adequacy checking. However, graphical diagnostics based on conventional residuals may have problems for sparse data as were pointed out by Dunn and Smyth (1996). They proposed randomized quantile residuals to make the aggregate pattern of the residuals become apparent. In this study, we further consider for sparse data with overdispersion which are frequently happened for Poisson and binomial type data. Simulations are conducted to examine the performance of the proposed modeling approach. In addition, we adaptively use the added variable plot and the partial residual plot based on randomized quantile residuals and exam in the relative performance via simulations. 丘政民 2000 學位論文 ; thesis 0 en_US |
| collection |
NDLTD |
| language |
en_US |
| format |
Others
|
| sources |
NDLTD |
| description |
碩士 === 國立中正大學 === 數理統計研究所 === 88 === Residuals play an important role in model adequacy checking. However,
graphical diagnostics based on conventional residuals may have problems
for sparse data as were pointed out by Dunn and Smyth (1996).
They proposed randomized quantile residuals
to make the aggregate pattern of the residuals become apparent.
In this study, we further consider for sparse data with overdispersion
which are frequently happened for Poisson and binomial type data.
Simulations are conducted to examine the performance of the proposed modeling
approach. In addition, we adaptively use the added variable plot and
the partial residual plot based on randomized quantile residuals and
exam in the relative performance via simulations.
|
| author2 |
丘政民 |
| author_facet |
丘政民 郭籠慶 |
| author |
郭籠慶 |
| spellingShingle |
郭籠慶 Some extensions to the use of randomized quantile residuals in generalized linear models |
| author_sort |
郭籠慶 |
| title |
Some extensions to the use of randomized quantile residuals in generalized linear models |
| title_short |
Some extensions to the use of randomized quantile residuals in generalized linear models |
| title_full |
Some extensions to the use of randomized quantile residuals in generalized linear models |
| title_fullStr |
Some extensions to the use of randomized quantile residuals in generalized linear models |
| title_full_unstemmed |
Some extensions to the use of randomized quantile residuals in generalized linear models |
| title_sort |
some extensions to the use of randomized quantile residuals in generalized linear models |
| publishDate |
2000 |
| url |
http://ndltd.ncl.edu.tw/handle/20142069927919244068 |
| work_keys_str_mv |
AT guōlóngqìng someextensionstotheuseofrandomizedquantileresidualsingeneralizedlinearmodels AT guōlóngqìng suíjīfēnwèishùcánchàzhíyòngyúguǎngyìxiànxìngmóxíngzhīyīxiētuīguǎng |
| _version_ |
1716849169153392640 |