-
- Nishina Kazuya
- National Institute for Environmental Studies
-
- Ishii Yumiko
- National Institute for Environmental Studies
Bibliographic Information
- Other Title
-
- 土壌動物学で利用できるRを使った統計手法の紹介 : 個体数や在/不在データに関わるエトセトラ
- ドジョウ ドウブツガク デ リヨウ デキル R オ ツカッタ トウケイ シュホウ ノ ショウカイ : コタイスウ ヤ ザイ/フザイ データ ニ カカワル エトセトラ
Search this article
Abstract
The article introduces the use of a basic statistical technique, Generalized Linear Models (GLMs), for dealing with both numerical population data and presence-absence data to soil ecologists. These models have recently become more popular for use with ecological studies and in soil zoology. This paper uses simple hypothetical examples and some actual datasets with a sample of R programming code in a way that allows the reader to study and learn these statistical analysis techniques through hands-on methods. We also provide various examples including figures related to the observational data, because effective data visualization is the most important part of statistical analysis. The first section introduces GLM particularly as it relates to Poisson and binomial regression analysis. The second section presents an example of the use of a Species Distribution Model (SDM), which allows the prediction of the potential spatial distribution of a species. SDMs using presence data of Yamatentomon yamato (Order Protura) across Japan were constructed by two different methods: GLM and random forest, the latter being a type of machine learning. Finally, the performances of the two models are compared.
Journal
-
- Edaphologia
-
Edaphologia 97 (0), 1-18, 2015
The Japanese Society of Soil Zoology
- Tweet
Keywords
Details 詳細情報について
-
- CRID
- 1390001204320515584
-
- NII Article ID
- 110010015637
-
- NII Book ID
- AN00334874
-
- ISSN
- 21898499
- 03891445
-
- NDL BIB ID
- 027013187
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- IRDB
- NDL
- CiNii Articles
-
- Abstract License Flag
- Disallowed