The agricultural segment is a major supporter of the Indian economy as it represents 18% of India's GDP, and it gives work to half of the nation's work power. The farming segment is required to satisfy the expanding need for food because of the increasing populace. Therefore, to cater to the ever-increasing needs of people of the nation, yield prediction is done prior. The farmers are also benefited from yield prediction as it will assist the farmers to predict the yield of crops prior to cultivating. There are multiple parameters that affect the yield of crops like rainfall, temperature, fertilizers, pH level, and other atmospheric circumstances. Thus, considering these factors, the yield of a crop is thus hard to predict and becomes a challenging task. In this chapter, the dataset of different states producing different crops in different seasons is considered; further, after preprocessing the data, the authors applied machine learning algorithms, and their results are compared.
TopIntroduction
India is a horticultural nation, with the majority of its population reliant on the industry. It is one of the most prevalent occupations in our country and contributes significantly to the country's overall development. Due to the reality that it's far the spine of the Indian economy. Around 60% of land is devoted to agriculture so that it will meet the needs of 1.2 billion people (Khan et al., 2021). Thus, modernization in horticulture is crucial and could advantage our nation's ranchers. The level of manufacturing in India is lowering due to different factors consisting of environmental changes, choppy rainfall, water management, and immoderate pesticide use. For a whole lot of reasons, the bulk of farmers fall quick of the anticipated crop yield. To comprehend production levels, crop yield prediction is used, which incorporates crop yield predictions into the data. Previously, crop yield estimation was highly dependent on the type of crop and the farmer's cultivation understanding.
There are several methods to address improvements and growth harvest yield and quality. Data mining strategies also are useful for forecasting crop yields. Generally, Data Mining is used to take a look at records from numerous frameworks and summarise it as beneficial records. Data mining software is a valid device that permits customers to explain and summarise obvious relationships, in addition, in addition, to take taking a look at records at numerous estimations or edges. Indeed, Data Mining is the technique of figuring out institutions or examples of fields (Priya et al., 2018). Data may be used to offer context for connections, institutions, or fashions. Then, this understanding may be converted into recorded fashions and records that may be used to create destiny examples or fashions. For instance, an overview of agricultural topics implores ranchers to indicate and avoid destiny harvest incidents.
Numerous analysts have been directed to develop a robust strategy for forecasting yields. However, the centre has consistently focused on quantifiable strategies and a small amount of work that has been prepared using a machine learning approach. Crop production is dependent on a variety of different parameters (Chowdary & Venkataramana, 2018), which vary according to climatic conditions such as temperature, humidity, rainfall, and soil pH value, as well as region geography and fertilisers. Numerous prediction tools for various crops make use of subsets of these parameters. Thus, prediction models are classified into two broad categories. There are statistical models that employ a single prediction task that encompasses all possible example spaces. Another approach is machine learning, in which information for data searches is used to connect the input and output variables.
AI can acquire capability in conjunction with the machine without requiring knowledge of PC programming; as a result, it advances machine completion by recognising and depicting the stability and example of driven data. We used supervised learning techniques to forecast crop yields in this study. For precise and effective methods, learning data is used to associate desired outputs with an input-output mapping standard. It entails developing a machine learning representation that is dependent on labelled sample.
Until now, prediction has been limited to a single state or crop. However, this proposed system investigates the application of supervised machine learning approaches in determining the yield production of a variety of crops in a variety of different states and their districts. This section collects, analyses, and tactics records from diverse states and crops. For yield estimation, techniques which include Linear Regression, Random Forest Regression, Gradient Boosting Regression, Polynomial Regression, Decision Tree Regression, and Ridge Regression had been used.
This chapter is organized in the following manner: Section II discusses related work in the fields of data mining and machine learning. Segment III consists of an in-depth discussion of the proposed framework. Segment IV will refer to the communiqué and its outcome. Finally, Section V completes the examination work.