Global Journal of Science Frontier Research, A: Physics and Space Science, Volume 23 Issue 1

about new data. Training ML models with larger volumes, more relevant, and accurate data increases the forecasting accuracy. Cognitive analytics helps analyse the various changes in a parameter, its effect, and its solution (Asakiewicz, 2016). Extensive studies on the predictive ability of ML techniques, such as multiple linear regression, regression trees, artificial neural networks, support vector regression, and k-nearest neighbor for crop yield production, have been undertaken (Gonzalez-Sanchez, 2014) . ML and advanced sensing techniques have been used to analyze online multi-layer soil data, and satellite imagery of crop growth characteristics to predict wheat yield (Pantazi et al, 2016) . ‘Crop Advisor’ is a software tool for predicting the influence of climatic parameters on crop yields (Veenadhari et al, 2014; Majumdar et al,2017) . SMAG, a French company has developed software using 30 years of weather data history, satellite and drone images, and soil types to enable users to track the progress during the life cycle of a plant and predict yields. 80% of French agricultural land under wheat cultivation is managed through this algorithm. InVivo, France’s leading agricultural cooperative group with 220 members using the technology, earns €6.4 billion in sales (Talend, n.d.) . II. K nowledge A griculture TM Knowledge Agriculture aims to achieve the objectives enshrined in SDG2030 to increase the agriculture yield without harming the environment, sustaining planet resources, and simultaneously countering challenges such as depletion of water resources and increased erosion and loss of productivity due to the occurrence of extreme weather events. The recent global environmental changes are apparent caution to mankind to immediately change the way it produces food and makes it available to the present and future generations. This can be achieved by optimal use of input resources such as seed, water, fertilizer and chemicals by preventing loss in storage, transportation and supply. In knowledge agriculture, data is captured from several sources. Sensors and IOT (Internet of Things) are used to collect farm data such as soil moisture, humidity, temperature, heat, light, etc. A second database stores the survey report that includes, productivity of soils, suitability of soils for raising specific crops, etc. Another database contains details of time for planting, produce to be planted, row spacing, desired yield, waste recycling, water supply, etc. There are several other databases which store historical rates, market details, supply chains, etc. These structured, semi-structured, or unstructured data are ingested into local or cloud-based storage, computing, analyzing, and creating models for ML-based predictions and taking corrective actions. The data allows the farmers to use a DSS (Decision Support System), to make optimized decisions such as tending each plant with water and fertilizers, spraying pesticides, and eliminating the weeds by selective burning or tilling as per its need during the life cycle without damaging the desired plant. Further, supply chain network design, product design and development, demand planning, procurement management, customized production, inventory management, logistics, and agile supply chain are essential components in farm management. A sustainable global supply chain is becoming an ever- increasingly complex system, especially with the need to deal with international partners (McCue, 2020). Modern technologies, such as CEP(Complex Event Processing), RFID, block chain, IoT, and WSN(Wireless Sensor Networks), boost supply chain performance. Supply chain analytics allows quick adjustments and more effective tactical decisions. III. S torytelling from F arm D ata a) System Architecture for the use of machine learning models for Knowledge Agriculture A system architecture for predictive analysis of historical and live data and taking corrective actions is shown in Figure 1. © 2023 Global Journals 1 Year 2023 18 Global Journal of Science Frontier Research Volume XXIII Issue ersion I VI ( A ) Data-Driven Knowledge Agriculture: A Paradigm Shift for Enhancing Farm Productivity & Global Food Security