查詢結果分析
來源資料
相關文獻
- 運用羅吉斯迴歸分析山坡地土地可利用限度查定分類
- 盤中檢查點的設計--簡單羅吉斯模型(Logistic Regression Model)的運用
- Tests for Response Probability in Multiple Logistic Regression Models
- 壽險業喪失清償能力信賴區間之研究
- 波段轉折分辨的技術--羅吉斯迴歸(Logistic Regression)分析的應用(以RSI為例)
- 設限資料下長期存活率之羅吉斯迴歸分析
- 五專推甄生與一般生在學期成績之比較--以臺北商專會統科為例
- 都會區工廠附近居民對汽機車排放污染物之感受程度與影響因子探討
- 應用二元羅吉斯迴歸模式及交叉效度法於我國壽險業務員雇用分界點決定之研究
- 企業對獎勵旅遊需求之研究
頁籤選單縮合
| 題 名 | 運用羅吉斯迴歸分析山坡地土地可利用限度查定分類=Classification of Slopeland Utilization Limitations by Logistic Regression Analysis |
|---|---|
| 作 者 | 林俐玲; 王兆文; 沈哲緯; | 書刊名 | 水土保持學報 |
| 卷 期 | 43:3 2011.09[民100.09] |
| 頁 次 | 頁277-296 |
| 分類號 | 434.278 |
| 關鍵詞 | 山坡地土地可利用限度查定分類; 羅吉斯迴歸; Classification of slopeland utilization limitations; Logisitic regression; |
| 語 文 | 中文(Chinese) |
| 中文摘要 | 查定工作係依平均坡度、土壤有效深度、土壤沖蝕程度及母岩性質等4 項因子,作為土地 類別等級之依據,惟查定人員往往現場判定受於地形複雜度而影響判定之準確性,故本研究期 以客觀且量化評估山坡地土地可利用限度查定分類之適宜性。 本研究以南投縣竹山鎮大坑段、大鞍段、筍子林、鯉魚尾及雲林縣古坑鄉草嶺段等5 個地 段為區域,5 個地段需辦理查定之地籍坵塊共計10,085 筆,其中宜農牧地計6,799 筆、宜林地 計2,915 筆及加強保育地計357 筆,透過SPSS 統計軟體亂數選取宜農牧地及宜林地各1,175 筆土地為訓練資料,餘為測試資料之使用。再運用羅吉斯迴歸方法,選取坡度、地形曲率、降 雨沖蝕指數、土壤沖蝕性指數、土地覆蓋與管理指數、土壤有效深度、坡地岩體強度、常態化 差異植生指數等因子進行山坡地土地可利用限度查定分類分析得知,宜林地與宜農牧地準確率 皆有七成以上,總體準確度達73%,而驗證總準確度亦達七成以上,模式預測成效尚稱良好, 故本研究建立之山坡地土地可利用限度查定分類模式,應可輔助查定人員現場判定參考,以提 升查定人員之行政效率。 |
| 英文摘要 | The verification tasks are categorized into classes according to factors including average slope, soil depth, soil erosion and parental rock. However, the accuracy of judging in the field made by the manual verification may be affected by terrain complexity; this study therefore employs quantified objective data for assessing the suitability of the classification of slopeland utilization limitations. The study takes 5 sections as the study area, which are Da-Keng Section, Da-An Section, Sun-Zih-Lin Section and Li-Yu-Wei Section in Jhushan Township of Nantou County and Tsao-Ling Section in Gukeng Township of Yunlin County. A total of 10,085 cadastral entries of hilly parcels are included in the 5 sections under study, of which 6,799 are suitable for agriculture or animal husbandry, 2,915 are suitable for forestry, and 357 are classified as requiring enhanced preservation. Through random selection, 1,175 entries each are chosen as training data by the SPSS software from the slopelands suitable for forestry and from slopelands suitable for agriculture or animal husbandry; the rest are used as test data. Followed by logistic regression analysis, factors including slope, terrain curvature, erosivity index, erodibility index, land cover and management index, soil depth, strength of rock mass, and normalized difference vegetation index are selected for classifying slopeland utilization limitations. The results show an accuracy of more than 70% with an overall accuracy of 73%. Total accuracy of the verification also exceeds 70%, the modal forecast is fairly good. Therefore the modal classification of slopeland utilization limitations built by the study shall be very useful to judging in the field made by the verifications in escalating his/her administrative efficiency. |
本系統中英文摘要資訊取自各篇刊載內容。