03327ntm a22005417i 4500 000151447 CZ-PrVSE 20220625121058.0 m d cr n|||||||||| 220625s2022 xr fsbm 000 0 eng d NEZPRACOVANÝ IMPORT ABA006 cze ABA006 ABA006 rda Kuzdas, Vojtěch ISIS:149089 dis Předpověď výnosu pšenice pomocí regresního přístupu eng Wheat yield prediction using a regression approach / Vojtěch Kuzdas 2022 ?? stran : digital, PDF soubor Vedoucí práce: Lubomír Štěpánek Bakalářská práce (Bc.)—Vysoká škola ekonomická v Praze. Fakulta informatiky a statistiky, 2022 Obsahuje bibliografii Textový (vysokoškolská kvalifikační práce) Rok obhajoby 2022 This bachelor thesis explores the applicability of linear and tree regression models in predicting the true yield of wheat crops in the South Moravian and Ústí nad Labem regions in the Czech Republic. Thesis pairs Sentinel-2 satellite imagery with single-season post-harvest yield data into a 10x10m grid. Yield data are split into percentile and manually filtered datasets. Two vegetation indices, NDVI and EVI, are extracted from satellite imagery and used as the primary independent variables in model training. Some models use additional post-harvest variables as independent variables. Simple and multiple linear regression and tree regression with bagging are done to predict the true yield. All models across all datasets are compared using standard R2 and MSE statistics. Models based on heavily filtered datasets achieve R2 values in the upper range of 0.92. Models based on less filtered datasets achieve R2 values of 0.65. However, the industrial applicability of constructed models requires further discussion due to potential issues (e.g. overfitting). Tree-based models showed better performance over linear regression models. Although perceived promisingly, tree-bagging did not prove to be effective on the given data. Způsob přístupu: Internet aplikovaná informatika [obor bakal. práce] bakalářské práce fd132403 czenas bachelor's theses eczenas Yield Prediction Remote Sensing Precision Agriculture Linear Regression Regression Trees Bagging Štěpánek, Lubomír ISIS:120809 ths Pinkas, Jaroslav opn Vysoká škola ekonomická v Praze. Fakulta informatiky a statistiky kn20010709399 dgg https://insis.vse.cz/zp/79596/podrobnosti VŠKP v InSIS https://insis.vse.cz/zp/79596 Hlavní práce https://insis.vse.cz/zp/79596/posudek/vedouci Hodnocení vedoucího https://insis.vse.cz/zp/79596/posudek/oponent/74667 Oponentura https://insis.vse.cz/zp/79596/priloha/23984 Přiloha k práci https://insis.vse.cz/zp/79596/podrobnosti dc:identifier NEPOSILAT VSKP vse79596 220625 79596