Quadratic discriminant feature selected broken stick regressive deep convolution neural learning classification for turmeric crop yield prediction

Agriculture holds a predominant position in the growth of any country's prosperity. However, there exists a major threat in the crop yield due to unpredictable and uncontrolled climatic changes, traditional farming methods, and poor irrigation services. In the past, crop yield was predicted based on farmer's experience. At present, the challenge is to increase the crop yield to meet the needs of the growing population. Currently, Machine Learning (ML) techniques are being used in multiple fields to accomplish practical and productive solutions. There are various algorithms in ML based on classification, clustering, and neural networks which can be used to predict the crop yield. In this work, we propose a method based on K-Nearest Neighbors (KNN) algorithm which detects the soil quality and predicts the suitable crop for cultivation. We consider temperature and soil quality as inputs to our algorithm. In addition, our method suggests the fertilizer based on the crop predicted. The test results show that our method accurately predicts the crop selection and yield which helps the farmers to great extent.

PurposeMany studies have proposed predictive models for type 2 diabetes mellitus (T2DM). However, these predictive models have several limitations, such as user convenience and reproducibility. The purpose of this study was to develop a T2DM predictive model using electronic medical records (EMRs) and machine learning and to compare the performance of this model with traditional statistical methods.Materials and MethodsIn this study, a total of available 8454 patients who had no history of diabetes and were treated at the cardiovascular center of Korea University Guro Hospital were enrolled. All subjects completed 5 years of follow up. The prevalence of T2DM during follow up was 4.78% (404/8454). A total of 28 variables were extracted from the EMRs. In order to verify the cross-validation test according to the prediction model, logistic regression (LR), linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), and K-nearest neighbor (KNN) algorithm models were generated. The LR model was considered as the existing statistical analysis method.ResultsAll predictive models maintained a change within the standard deviation of area under the curve (AUC) <0.01 in the analysis after a 10-fold cross-validation test. Among all predictive models, the LR learning model showed the highest prediction performance, with an AUC of 0.78. However, compared to the LR model, the LDA, QDA, and KNN models did not show a statistically significant difference.ConclusionWe successfully developed and verified a T2DM prediction system using machine learning and an EMR database, and it predicted the 5-year occurrence of T2DM similarly to with a traditional prediction model. In further study, it is necessary to apply and verify the prediction model through clinical research.

Agriculture plays a crucial role in the economic development of many countries, especially where farming is the primary source of livelihood. Farmers often face difficulties in selecting suitable crops, predicting crop yield, identifying plant diseases at an early stage, and estimating financial outcomes such as profit or loss. To address these challenges, this study presents Sheti Mitra, an interactive machine learning–based system designed to assist farmers in making informed agricultural decisions. The proposed system integrates crop recommendation, crop yield prediction, and plant disease detection into a single platform. The system utilizes image processing and deep learning techniques to detect diseases from leaf images using Convolutional Neural Networks (CNN), while crop recommendation and yield prediction are performed using machine learning algorithms such as Random Forest. Environmental parameters including soil nutrients, temperature, humidity, and rainfall are analyzed to determine the most suitable crop and estimate expected yield. Additionally, the system provides an estimation of profit or loss based on predicted yield and crop market factors, helping farmers plan their agricultural activities effectively. A web-based interface built using HTML, CSS, and JavaScript interacts with a Flask-based backend that processes user inputs and generates predictions using trained models. By combining disease detection with crop and yield prediction, the proposed system offers a comprehensive decision-support tool for farmers. The implementation demonstrates that machine learning techniques can significantly improve agricultural productivity and reduce risks associated with crop selection and disease management. This system aims to support smart farming practices and promote sustainable agricultural development.

The fuzzy quadratic classifier (FQC) extends the quadratic classifier to accommodate a sequence of fuzzy numbers as data. The fuzzy data induces a fuzzy quadratic discriminant (QD) that is then compared to other fuzzy discriminants to produce a decision. To generate the fuzzy data, conversion of batches of data vectors, or signals, into sequences of fuzzy signal vectors is developed base on fuzzy order statistics. In data rich environment, this conversion both condenses and refines the data stream. An efficient implementation of the FQC is presented based upon a diagonal form of the quadratic classifier and robust estimation of the parameters. A two class problem illustrates the procedure.

Crop yield forecasting plays a crucial role in farming planning and management, international food trade, ecosystem sustainability, and so on. There are many factors that have an influence on crop yield, such as plantation area, efficiency of irrigation systems, variations in rainfall and temperature, quality of crop seeds, topographic attributes, soil quality and fertilisation, and disease occurrence. In the recent years, many researches support that data mining techniques algorithms can be profitable applied to a wide range of agriculture problems and, in particular, to have reasonable crop yield predictions. The aim of this chapter to compare the potentiality of two hybrid spatial interpolation techniques for the spatial estimation of the crop yield: regression kriging (RK) and random forest (RF). As a case study, we interpolated the winter wheat yields in the Southern U.S. Great Plains.

Accurate prediction of food crop yield is of great significance for global food security and regional trade stability. Since remote sensing data collected from unmanned aerial vehicle (UAV) platforms have the features of flexibility and high resolution, these data can be used as samples to develop regional regression models for accurate prediction of crop yield at a field scale. The primary objective of this study was to construct regional prediction models for winter wheat yield based on multi-spectral UAV data and machine learning methods. Six machine learning methods including Gaussian process regression (GPR), support vector machine regression (SVR) and random forest regression (RFR) were used for the construction of the yield prediction models. Ten vegetation indices (VIs) extracted from canopy spectral images of winter wheat acquired from a multi-spectral UAV at five key growth stages in Xuzhou City, Jiangsu Province, China in 2021 were selected as the variables of the models. In addition, in situ measurements of wheat yield were obtained in a destructive sampling manner for prediction algorithm modeling and validation. Prediction results of single growth stages showed that the optimal model was GPR constructed from extremely strong correlated VIs (ESCVIs) at the filling stage (R2 = 0.87, RMSE = 49.22 g/m2, MAE = 42.74 g/m2). The results of multiple stages showed GPR achieved the highest accuracy (R2 = 0.88, RMSE = 49.18 g/m2, MAE = 42.57 g/m2) when the ESCVIs of the flowering and filling stages were used. Larger sampling plots were adopted to verify the accuracy of yield prediction; the results indicated that the GPR model has strong adaptability at different scales. These findings suggest that using machine learning methods and multi-spectral UAV data can accurately predict crop yield at the field scale and deliver a valuable application reference for farm-scale field crop management.

Assessing crop growth during the growing period and the prediction of crop productivity is gaining importance for the estimation of seasonal production prior to crop harvest. Reliable and timely estimates of crop production in the face of different types of abiotic or biotic stresses or climatic aberrations, namely heat stress, drought or pest infestation, etc., are particularly important for those countries where agriculture contributes substantially to the economy. Advance warnings of adverse effects of such disastrous events can be of great help for strategic planning and the use of measures to avoid any consequent food shortages in the future. Satellite remote-sensing platforms can be applied quite effectively for real-time crop condition assessment and yield predictions, as satellite systems can provide continuous data on variable spatio-temporal scales, with timely availability and global coverage. These data are also inexpensive and readily available at various online portals and archives. Several studies conducted so far have either derived vegetation indices from remote-sensing data or have coupled remote-sensing data with crop simulation models for crop yield prediction. The coupling of remote-sensing data with crop simulation models has shown a significant improvement in yield prediction. Among spectral indices, Normalized Difference Vegetation Index (NDVI) has been shown to be quite effective for predicting crop yield, but detection of the anomalies from historical values of NDVI needs the availability of NDVI at the temporal scale for predicting the crop yield of the current season. Obtaining biophysical parameters (leaf area index, soil moisture, evapotranspiration, etc.), derived from satellite imagery at the appropriate stage of crop growth, is required for robust prediction of crop yield at the regional level. Yield estimates at regional level can provide important information for agricultural planning and can enable policy makers to frame import or export plans, determining prices on one hand, and can enable farmers to plan their marketing strategies on the other hand, while ensuring food security.

Developing machine learning models with multi-source environmental data to predict wheat yield in China

Rice is one of the six priority crops chosen by the government of Rwanda to increase agricultural productivity. This crop faces significant challenges from climate variability and soil conditions, affecting yields and food security. The past studies concentrated on building machine learning regression models for predicting the rice productivity with the help of weather parameters as predictors. Few researchers incorporated some soil nutrients to improve the precision of the prediction model. However, most of the farmers in Rwanda rely on agricultural input like inorganic fertilizer to boost the agriculture productivity. This means that the weather parameters are not sufficient to build a robust crop yield prediction model without incorporating the soil parameters. In this paper, rice yield status (favorable, unfavorable) prediction model was optimized using machine learning techniques, including logistic regression, random forest, gradient boosting, extreme gradient boosting, support vector machine, and artificial neural networks. The models were trained on historical weather parameters and soil nutrient data collected from the National Institute of Statistics of Rwanda, Rwanda Agriculture and Animal Resources Development Board and Rwanda Meteorology Agency. Findings indicated that the eXtreme Gradient Boosting (XGBOOST) model outperformed other models under consideration with 91%, 94.1%, and 82.7% of prediction accuracy, sensitivity, and precision, respectively. The above results demonstrate that the XGBOOST is the most robust model in predicting rice yield. The feature importance technique applied in this study also revealed that weather parameters (rainfall, and temperature), phosphorus, sulfur, manganese and carbon were crucial factors to prioritize for rice yield prediction compared to other model predictors under study. The results may provide valuable insights into whether the rice yield will be classified as low or high. This information is essential for helping farmers and decision-makers obtain actionable insights into expected crop yields, enabling them to make informed decisions on agricultural inputs resource allocation to maximize Rwanda's agricultural productivity. Key words: artificial intelligence, machine learning, precision farming, Smart agriculture, yield prediction

&lt;abstract&gt; &lt;p&gt;It is becoming increasingly evident that educators need to prioritize the welfare of their students, particularly those who are underperforming academically, also known as "students at risk". By analyzing learning behaviors, including attendance records, past academic results, and online interactions, we can identify students at risk and provide them with timely support. Therefore, this study aimed to develop a prediction model for identifying students at risk in an actuarial science course and suggest an intervention strategy. Our study was comprised of five components of learning analytics: data collection, reporting, prediction, intervention, and reassessment. Prior to applying a prediction model, correlation analysis was utilized to identify variables impacting students' academic performance. Three variables, including CGPA, pre-requisite subject marks, and assessment marks were considered due to their rather strong correlation with the final marks of the course. Then, quadratic discriminant analysis (QDA) was applied to predict students classified as "at risk" and "not as risk". Out of 69 students from the course, 15 students identified as "at risk" and 40 students participated in the Peer Assisted Learning Program (PALP) as an intervention strategy to reduce the course's failure rate. We cannot conclude whether PALP was an effective intervention strategy for students at risk because a majority of them failed to attend. However, we observed that those who attended PALP had a higher likelihood of passing the course. Our prediction model had high rates of accuracy, precision, sensitivity, and specificity, which were 91%, 98%, 91%, and 91%, respectively. Therefore, QDA could be considered a robust model for predicting students at risk. We have outlined some limitations and future studies at the end of our study.&lt;/p&gt; &lt;/abstract&gt;

In this innovative project, we aim to revolutionize agriculture by harnessing advanced technology to empower farmers with predictive insights into crop yields even before the planting season begins. Our approach combines the analysis of soil data and images of the soil to develop a robust predictive model. By leveraging machine learning algorithms, we can decipher intricate patterns within the soil data and images, enabling accurate predictions of crop growth potential based on soil characteristics. This groundbreaking technology holds immense promise for optimizing farming practices, allowing farmers to make informed decisions regarding crop selection and yield expectations, ultimately leading to increased efficiency and productivity in food production. The foundation of our project lies in the comprehensive analysis of soil data, encompassing crucial parameters such as nutrient composition and moisture levels. Soil health plays a pivotal role in determining crop growth and yield, making it imperative to understand its characteristics thoroughly. Through advanced sensing technologies and data collection methods, we gather detailed information about the soil, providing a comprehensive dataset for analysis. By scrutinizing this data, our machine learning algorithms can discern meaningful patterns and correlations, thereby gaining insights into the soil's suitability for different crops. These images offer a visual representation of the soil's physical attributes, such as texture, structure, and moisture distribution. By integrating image analysis with soil data, we enrich the dataset and enhance the depth of our predictive model. Machine learning algorithms trained on this combined dataset can effectively interpret visual cues and extract valuable insights, contributing to more accurate predictions of crop yield potential.

Agrometeorological systems for regional crop yield forecasting have traditionally relied on weather data derived from weather stations for crop simulation and yield prediction. In recent years, numerical weather prediction (NWP) models have become an interesting source of weather data with the potential to replace observed weather data. This is a result of the steadily decreasing NWP grid sizes and the availability of long and consistent time-series through the so-called reanalysis projects. We evaluated the ERA-INTERIM reanalysis data set from the European Centre for Medium-range Weather Forecasting for regional crop yield forecasting. Crop simulations were carried out using 2 identical model implementations: one using interpolated observed weather, the other using weather data derived from ERA-INTERIM. Output for both sources of weather variables was generated for the EU27 and neighbouring countries and 14 crops, aggregated to national level and validated using reported crop yields from the European Statistical Office. The results indicate that the system performs very similar in terms of crop yield forecasting skill for both sources of weather variables. In 38% of the crop–country combinations, the forecasting error can be reduced by more than 10% of the baseline forecast (the trend only) for both sources of weather variables. In almost 20% of the crop–country combinations, the forecasting error can be reduced by more than 25% of the baseline forecast. The results demonstrate that the ERA-INTERIM data set is highly suitable for regional crop yield forecasting over Europe and may be used for implementing regional crop forecasting over data-sparse regions. Finally, we conclude that there is a need to improve the crop calendar and/or calibration for some of the modelled crops.

Climate variability poses a significant threat to global agricultural productivity and food security, particularly in regions that are highly dependent on climate-sensitive farming systems. Accurate and timely prediction of crop yields is therefore essential for informed decision-making and policy formulation. This study presents an AI-driven predictive modeling framework for estimating crop yields under varying climatic conditions using historical agricultural and climate data. A publicly available global crop yield dataset incorporating rainfall, temperature, and pesticide usage was utilized to develop and evaluate predictive models. The traditional statistical regression was compared to the machine learning approaches, namely Random Forest and Gradient Boosting regressors, to determine the predictive performance and strength of the models. The standard regression metrics, which included root mean square error (RMSE), mean absolute error (MAE), and the coefficient of determination (R²), were used for model evaluation. The findings of the study showed that the models based on machine learning methods completely outclass the traditional linear regression in grasping the nonlinear relations between climate factors and crop yield. The analysis of feature importance also points out specific regional and temporal variations in rainfall and temperature as the two leading contributors to changes in crop yield fluctuations, thus implying the vulnerability of agricultural productivity to the effect of climate change. The results highlight the ability of models predicting crop yield using AI technology to be one of the main supports for climate-sensitive agricultural planning and global food security improvements. This work offers one more example of AI being used in the agricultural sector and the policy area, also as it is an input to making climate change and its effects on crop production data-driven matters, thus increasing the whole AI 'sustainable agriculture' and 'evidence-based policy area.

Accurate early season predictions of crop yield at the within-field scale can be used to address a range of crop production, management, and precision agricultural challenges. While the remote sensing of within-field insights has been a research goal for many years, it is only recently that observations with the required spatio-temporal resolutions, together with efficient assimilation methods to integrate these into modeling frameworks, have become available to advance yield prediction efforts. Here we explore a yield prediction approach that combines daily high-resolution CubeSat imagery with the APSIM crop model. The approach employs APSIM to train a linear regression that relates simulated yield to simulated leaf area index (LAI). That relationship is then used to identify the optimal regression date at which the LAI provides the best prediction of yield: in this case, approximately 14 weeks prior to harvest. Instead of applying the regression on satellite imagery that is coincident, or closest to, the regression date, our method implements a particle filter that integrates CubeSat-based LAI into APSIM to provide end-of-season high-resolution (3 m) yield maps weeks before the optimal regression date. The approach is demonstrated on a rainfed maize field located in Nebraska, USA, where suitable collections of both imagery and in-situ data were available for assessment. The procedure does not require in-field data to calibrate the regression model, with results showing that even with a single assimilation step, it is possible to provide yield estimates with good accuracy up to 21 days before the optimal regression date. Yield spatial variability was reproduced reasonably well, with a strong correlation to independently collected measurements (R2 = 0.73 and rRMSE = 12%). When the field averaged yield was compared, our approach reduced yield prediction error from 1 Mg/ha (control case based on a calibrated APSIM model), to 0.5 Mg/ha (using satellite imagery alone), and then to 0.2 Mg/ha (results with assimilation up to three weeks prior to the optimal regression date). Such a capacity to provide spatially explicit yield predictions early in the season has considerable potential to enhance digital agricultural goals and improve end-of-season yield predictions.

Crop yield prediction is vital for global foodstuffs and intensely challenging due to its multitudinous factors such as soil, weather, rainfall, temperature, and their interactions. The influence of weather changes in India impacted in terms of the agricultural crop outcomes over the last decades. The proposed research focuses on the weather condition and assists the farmers to comprehend the crop yields before growing onto farming. This research illustrates a deep learning framework using artificial neural networks (ANN). The offered ANN model, along with further widespread approaches such as bagging classifier, extra tees classifier, linear discriminant analysis (LDA) classifier, quadratic discriminant analysis (QDA) classifier and stochastic gradient boosting (SGB) classifier are used for India's downloaded crop dataset. These proposed six models achieved quite good accuracies. The numerous substantial contributions of the prediction models are the ability to produce the exact predictions. These were succeeded by training the algorithms and naturally to the verify data. Eventually, we tend to compare our study with the previous study's research paper.