Article Preview
Top2. Literature Review
The different analytic techniques like Principal component analysis, correlation matrix, and regression analysis were used for the prediction of soil reaction (pH), soil nutrients and correlation with positive and negative signs in the soil. These resulted in suggestion of fertilizer to the right soil type, nutrient level and pH level in order to produce healthy crop yield (Swapna et al., 2020).
The study of soil temperature through physical and chemical process is essential for land-atmosphere interaction, which is done through four machine learning techniques namely, ELM, ANN, CART, GMDH. In the developed models' climatic variables are utilized. Effects in increase and decrease of the soil depth are observed and are mapped using air temperature data input. Solar radiation and wind speed are also considered. In the four above mentioned techniques, Extreme learning machine (ELM) is most preferred (Alizamir et al., 2020).
In the field of agriculture, remote sensing systems are widely used to improve yield production and nitrogen management. This helps reduce cost and environmental impact. These require remotely sensed data that were extracted from machine learning methods, as it can process a large number of inputs and handle non-linear tasks. At the outcome, we conclude that applications of sensor modalities and ML techniques help develop a hybrid system being a part of precision agriculture (PA) in the future (Chlingaryan et al., 2018).Through geotechnical engineering, it has been established that using machine learning techniques helps functional correlations. This correlation has been ranked based on coefficient determination and absolute development. This applies to individual performances also. The results indicate that developed models are accurate and can provide a viable tool for engineers for the prediction of observed values (Puri et al., 2018).
The model predictive control method plays a vital role in agriculture as it effectively addresses non-linear and large time-delay systems. These are divided into three stages that support irrigation systems, agricultural machinery, product processing, and greenhouses without many changes in agricultural production regulation methods. The increasing present-day population is backed up by this method were water shortages, limited land resources, and low production efficiency can be regulated (Ding et al., 2018).In this paper, usage of official Spanish phytosanitary products registry, applying the combinations of the machine learning algorithm, natural language technique, and resampling techniques are best performing approaches through syntactical patterns. The main aim of this system is to incorporate the tractor-centric process that was defined for agricultural services (Espejo-Garcia et al., 2018).
India, an agricultural country having different climatic conditions. A machine learning technique comes handy to the farmers and other stakeholders to make decisions in agronomy and crop choice. This spotlights rice cropping areas, the precipitation, and the minimum, maximum temperature of the room. Through artificial intelligence, crop analysis is done, and farmers were provided with sufficient information about increasing their crop yield (Gandhi & Armstrong, 2016).
Using five functional parameters in different machine learning algorithms in soil properties prediction have resulted in indicating a location's ability to perform critical ecological services. The five soil characteristic properties such as Soil Organic carbon (SOC). Calcium, phosphorus, sand contents and pH value were accurately observed and enhanced using the machine learning techniques for better results (Akinola et al., 2018).The machine learning algorithms provides Digital soil mapping for the pedologist to predict the climate, lithology, vegetation and landforms. Application of artificial neural network and regression models are also taken into account. These enable comparison and identification of algorithms for specific soil=landscape conditions (Meier & de Souza, 2018).