引導式測驗 · 中等 等級

人工智慧在農作物產量預測測驗的應用

檢視植被指數、季節不確定性、堅持預測、衛星輸入、混合作物模型和產量驗證標籤的規模。

問題 1 的 8

NDVI植被指數主要反映什麼?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. NDVI植被指數主要反映什麼?
    • 土壤 pH 值
    • 糧食價格
    • 衛星光反射顯示農作物的綠色和活力
    • 農場工人數
  2. 為什麼早季產量預測比晚季產量預測有更多不確定性?
    • 影響產量的大部分天氣和作物生長尚未發生
    • 衛星在賽季初期關閉
    • 農民們早早就隱藏了自己的田地
    • NDVI 僅在冬季有效
  3. What does held-out-year validation help assess in a crop-yield forecast study?
    • A guarantee that no future information can leak into the inputs
    • Only whether the model's filename contains the year
    • Proof that every field shares the same yield
    • Generalization to another year's conditions, with forecast-time feature availability still needing control
  4. 哪個資料來源提供這些模型中常用的多光譜影像?
    • 限家庭氣象站
    • Sentinel-2 和 Landsat 衛星
    • 社群媒體貼文
    • 雜貨店收據
  5. What does the cited US Corn Belt hybrid yield study add to machine-learning predictors?
    • Variables supplied by a process-based crop simulation model such as APSIM
    • Removal of every weather and soil input
    • A guarantee that the model needs no regional validation
    • Only random labels unrelated to crop development
  6. Why should crop-yield evaluations examine unusual heat or drought years separately?
    • Extreme conditions always make prediction easier
    • A high NDVI guarantees an unchanged yield under every stress
    • Training mainly on ordinary conditions may not represent those stress responses adequately
    • Every weather year is identical once a crop is named
  7. If a vegetation observation has near-infrared reflectance 0.6 and red reflectance 0.2, what is its NDVI?
    • 0.8, by adding the two values only
    • 0.5, using (0.6 minus 0.2) divided by (0.6 plus 0.2)
    • 0.4, by subtracting without normalization
    • 3.0, by dividing near-infrared by red only
  8. What does validation against county-average yield labels establish about an individual-field forecast?
    • It guarantees every field has the county's average yield
    • It converts a regional label into a direct harvest measurement for each field
    • It removes all need to know the forecast's spatial unit
    • It does not by itself establish field-level accuracy; the intended scale needs its own evidence