Loading...
Thesis
Browse
22 results
Search Results
ThesisItem Open Access Evaluation of Artificial Neural Network (ANN) and Penalized Regression Model for Prediction of Rice and Maize Yield Based on Weather Parameters in Kashmir(SKUAST Kashmir, 2023) Bellikatti, Revanasiddappa Shekhappa; Bhagyashree DhekaleJammu and Kashmir (J&K) is an agrarian state, among all the agricultural states of India. J&K occupies 15th rank in term of agriculture. Rice is the major food crop for more than 90% of the Kashmir population. Rice and Maize is farmed on an area of 133.28 and 69.12 thousand hectares respectively, whereas it is produced in 3410 and 966 thousand quintals at a productivity of 25.59 and 13.98 quintal/ha respectively in the Kashmir division (DES, 2021). The effect of weather parameters plays an important role in agricultural production such as on plant growth, development, and harvesting. Climate change has affected the production of major crops in Kashmir region such as rice, maize, wheat, barley, pulses, and oilseeds, along with various fruits (Shabina Tabasum et.al., 2018). Changes in weather parameters do affect the production of Rice and Maize. Forecasting of yield of Rice and Maize is used to analyse past and present behaviours to predict a future production that does assist in decision-making and planning for the future effectively as well as efficiently. In the present analysis evaluation of artificial neural network (ANN) and penalized regression model for yield prediction of Rice and Maize based on the weather parameters in Kashmir has been used. The base of our study was the secondary data on the total yield of Rice and Maize crop all over Kashmir for 27 years from 1995-2021 (collected from www.indiaagristat.com) was undertaken. The meteorological data on four weather parameters (Max Temperature and Min Temperature (degrees in Celsius), and rainfall (mm) were collected from the Indian Meteorological Department (IMD). In this study, two models have been developed and evaluated for yield prediction of Rice and Maize, namely Artificial Neural Network (ANN) and Penalized Regression Model (Ridge regression and LASSO regression). Correlation and multicollinearity are worked out between the explanatory variables. Linear, ridge, lasso and ANN models were fitted to the rice and maize of production and yield. The performance of the forecasting model was tested using maximum R2, minimum RMSE, MAE, MAPE, AIC and BIC. In case of rice and maize, there exist significant correlation among the explanatory variables and there exists multicollinearity. Linear, Ridge, Lasso and ANN models were fitted and evaluated for its performance. For rice production and yield, among all the evaluated models ANN has the highest R2 (0.84 and 0.87) and lowest RMSE (288808.6 and 2.65), MAE (248816.9 and 2.55), MAPE (4.91 and 7.18), AIC (261.62 and -45.68) and BIC (274.38 and -32.11) during training and lowest RMSE (1151.95 and 3.68), MAE (248816.83 and 2.64) and MAPE (4.91 and 7.18) during validation. The forecasting is made for next six years using the best fitted model In case of Maize production and yield, among all the evaluated models Ridge regression has highest R2 (0.92 and 0.91) and lowest RMSE (293559.6and 2.51), MAE (250944.4 and 2.21), MAPE (42.57 and 14.64), AIC (272.84 and -2.04) and BIC (280.61 and -4.48) during training and lowest RMSE (1114.45 and 3.68), MAE (234277.72 and 2.54) and MAPE (40.25 and 15.93) during validation. The forecasting is made for next six years using the best fitted model.ThesisItem Open Access A Statistical Investigation on Association between Weather Parameters and Horticultural Crop Yield in Kashmir Region(SKUAST Kashmir, 2022) Omiya; Nageena NazirApple (MalusDomestica) is the most widely produced horticulture crop in the state of Jammu and Kashmir followed by Pear (Pyruscommunis) and Cherry(Prunusavium) . As per the horticulture census 2016-2017, about 48% of the area is covered under apple followed by pear which is 4.29% and cherry which is 0.83%. Among the three, apple is the most significant in terms of productivity, as it generates the greatest amount of marketable excess.Apple constitutes the largest production in UT of Jammu & Kashmir (86.55%) followed by pear (3.30%) and cherry (0.56%) respectively. An attempt has been made in the present study to conduct a statistical investigation on the association between weather parameters and horticultural crop yield in the Kashmir region, which was carried out for the period of 2001–2019. The statistical tools namely correlation analysis, regression analysis and different linear and non-linear models were employed. The results revealed that there is an association between weather parameters and crop yield. Among selected weather parameters maximum temperature is positively and minimum relative humidity is negatively correlated with crop yield. The models were created in order to predict yield using specific meteorological parameters. The best model was chosen based on the significance of variable and R2, which explains the variable in the dependent variable caused by the independent variable. Different non-linear models were used for predicting yield using each weather parameters, among those S-curve, Growth, Compound, Power, Linear, Quadratic model were found significant.ThesisItem Open Access Time Series Modeling and Forecasting of Area, Production and Productivity of Major Dry Fruit Crops of Kashmir(SKUAST Kashmir, 2022) Mehrukh Fatima; Bhagyashree DhekaleDry fruit crops are high-value, export-oriented commercial crops. In addition to their commercial significance, they also create a large number of employment opportunities. Dry fruit production in J&K was 275 thousand MT in 2016-17, occupying around 96 thousand hectares of land. Walnuts and almonds cover nearly all the dry fruit land in J&K UT, accounting for 98 percent of dry fruit production and12.19 percent of overall production. Long term time series data on Area, Production and Productivity of walnut and almond crop were collected from Directorate of Horticulture, Jammu and Kashmir. Descriptive statistics were required in order to obtain structural information of the data. The average of productivity of walnut was 2.47 MT/ha and for almond 0.91 MT/ha. Data was tested for randomness, normality and no outliers were present in time series data. To evaluate the trend in area, production and productivity of walnut and almond linear, quadratic and cubic models were fitted. The best fitted models were selected using the criteria of higher R2 and lower values of RMSE, MSE and MAPE. Based on the model adequacy, the quadratic model was the best fit for walnut area and productivity, whereas for production cubic model was found to be the best fit. In almond, cubic model was shown to be the best fit for trends in area, production and productivity. To forecast the dry fruit production, ARIMA model were used. As a series was found non stationary using ADF test, stationary conditions were achieved using differencing one. Based on R2, RMSE, MAPE, MAE, Max APE, Max AE and BIC, ARIMA (1,1,1), ARIMA (3,1,2) and ARIMA (3,1,2) were found to be the best fitted for area, production and productivity of walnut. Whereas, ARIMA(3,1,2), ARIMA(2,1,3) and ARIMA (2,1,2) were found to be most suitable models for area, production and productivity of almond. Selected models were validated for two years and found suitable for forecasting the area, production and productivity of walnut and almond crop. From the forecasting, it is observed that the area under walnut crop is decreasing while production and productivity is increasing in coming years. From the best fitted ARIMA models it is forecasted that almond’s area, production and productivity are decreasing.ThesisItem Open Access Instability and decomposition analysis of major Horticulture crops in Kashmir(SKUAST Kashmir, 2022) Arif Bashir; Khan, ImranThe goal of the current study was to look into Kashmir’s apple and pear crop production behaviour. The study used time-series data on the area, production, and productivity of the Apple and Pear crops for the years 2001-02 to 2019-20. The entire study was further split into two sub-periods, namely period I (2001-02 to 2009-10) and period II (2010-11 to 2019-20), and each sub-period was examined separately, district and zone wise. Instability and decomposition analysis was conducted in relation to the area, production, and productivity for the specified crop. Sen (1967) and Bandyopadhaya (1989) methodologies were used to determine the degree of instability in the production of apple crop, and its nature and effects were researched. In order to study changes in the average production and in the variance of production of Apple and Pear crop, a decomposition analysis was carried out as suggested by Hazell (1982, 1984). The examination of instability and decomposition’s findings showed that not all districts experienced stable apple and pear crop production. The degree of instability varied greatly between periods I and II. Large variations in mean production in both the crops were explained by components of changes in mean yields and mean areas. Changes in yield variance, area variance, and change in mean area were discovered to be the main causes of production instability in the majority of the districts.ThesisItem Open Access Design and Development in Ranked Set Sampling(SKUAST Kashmir, 2022) Bhat, Ishfaq Ahmad; Mir, S.A.Every scientific research area or anything dealing with the collection, processing, analysis, and interpretation of data is dependent on statistics. Therefore, the target of achieving the required results is based on a sound statistical platform assisting the researchers and experimenters to come up with better results and conclusions. In this regard, a new sampling strategy called as “Dual Ranked Set Sampling” (DuRSS) has been developed for its use in obtaining a valid sample from the population and also to obtain an efficient estimate of unknown population parameters. DuRSS helps in reducing the ranking error thereby increasing its efficiency. For situations where it is difficult or time-consuming or costly to identify m3 units from the target population, especially when there is a shortage of experimental units a new sampling strategy called as “Paired Systematic Ranked Set Sampling” (PSRSS) has been developed for its use in obtaining a valid sample from the population and also to obtain an efficient estimate of unknown population parameters. It uses comparatively less number of experimental units than the double ranked set sampling. In the final part a new sampling strategy based on neoteric ranked set sampling called “Neo-Centric Ranked Set sampling” (NCRSS) has been used to obtain an efficient estimate of unknown population parameters. The developed sampling schemes were subjected to preliminary study using real data set on high density apple plantation. Mean square error of the DuRSS, PSRSS and NCRSS estimator different sizes of m are computed and compared with the MSE of the estimators under SRS and RSS and DRSS to find their efficiency. Besides this a simulation study was carried out for different symmetric distributions using 1000 simulations for various set sizes and their relative efficiencies were calculated. The estimators of the developed sampling strategies out performed their counterparts under SRS, RSS and DRSS in all the cases. Empirical studies were carried out on real data using SAS and R. Various commands were executed using R platform.ThesisItem Open Access Classical and Ridge Regression Coefficient Estimators in presence of Multicollinearity in High Density Apple data(SKUAST Kashmir, 2021) Ume Kulsum; Khan, Dr. ImranThe agricultural investigations are based on the application of statistical methods and procedures which are helpful in testing hypotheses using observed data, in making estimations of parameters and in predictions. The application of statistical principles and methods is necessary for effective practice in resolving the different problems that arise in the many branches of agricultural activity. To find out one such suitable estimation procedures for Gala species of High Density apple, the present investigation has been carried out in detail. For detection of multicollinearity in the data, three different measures were used including examination of correlation matrix, variance inflation factors and eigenvalues inspection for various independent variables under study. The result revealed that problem of multicollinearity was serious in the study. The significant variables were selected using the all possible regression approach. Presence of influential observations were identified with the help of three different measures viz. Cook’s D Statistic, diagnosis of diagonal elements of hat matrix and DFFITS. In this investigation two estimation procedures, i.e., ordinary least squares and ridge regression technique were compared with each other. Mean square error was used to compare the performance of each estimator. The value of ridge constant (θ) was estimated by three different methods, viz. ridge trace technique, method given by Hoerl, Kennard and Baldwin and cross-validation method. All the methods of ridge regression were found to perform better than ordinary least squares method in terms of mean square error criterion. However, based on the minimum value of mean square error, ridge trace method was selected to calculate the optimum value of ridge constant. .ThesisItem Open Access A Study on Climate Change Trend and its Impact on Productivity of Paddy Crop under Temperate Conditions of Kashmir(SKUAST Kashmir, 2021) Rohit Godara; Sofi, Dr. Nazir AhmadThe present study attempts to know the trend of selected weather parameters and to analyze the impact of weather parameters on productivity of rice in Kashmir valley. The secondary data on weather parameter over period 1995-2019 was collected from, the Section of Agronomy, SKUAST-K Shalimar. Further, the secondary data of rice yield over period 1995-2019 was collected from, Directorate of Economics & Statistics, government of J&K. The parameters consider viz., average maximum temperature, average minimum temperature, average rainfall and rice yield for the study. The trend analysis revealed that the mean maximum temperature, mean minimum temperature and rainfall showed a positive insignificant trend over a period of time for most of the months. Also, the non-parametric method indicates that there is not a time dependent trend for all the months. The correlation analysis indicated that the productivity of rice was positive insignificant correlation with maximum temperature (0.271) and rainfall (0.374), whereas minimum temperature showing significant correlation (0.614). The minimum temperature contribution was higher (0.566 towards the productivity of rice, followed by maximum temperature (0.336) and rainfall (0.309).It was noticed that riced yield in Kashmir valley more dependent on minimum temperature.ThesisItem Open Access Statistical Modeling and Distribution Patterns Assessment of Major Conifers in Shopian Forest Division of Kashmir(SKUAST Kashmir, 2021) Aqib Gul; Bhat, Dr. Bilal AhmadThe present investigation entitled “Statistical Modeling and Distribution Patterns Assessment of Major Conifers in Shopian Forest Division of Kashmir” was carried out in Shopian Forest division. For the study, primary data for diameter at breast height and height of conifer plantations from two ranges viz. Shopian and Roamshi were collected. The data were subjected to variability analysis in order to check the variability between the forest ranges with respect to the growth parameters viz. diameter at breast height, height and volume and f-test was used for testing the homogeneity of variances. Various probability distributions were fitted to find out the expected number of trees in each diameter class and their significance was tested using Kolmogorov-Smirnov test statistic. Gamma distribution followed by lognormal distribution were observed best fitted for all the three species viz. Cedrus deodara, Abies pindrow and Pinus wallichiana, respectively for the estimation of number of trees in various diameter classes. Regression analysis was carried out to study the tree volume models and the best model was selected for volume estimation on the basis of maximum R² and R̄², minimum RMSE and Theil’s U statistic, whereas validation was tested by Chow test. In the case of Cedrus deodara and Pinus wallichiana, the cubic and quadratic model and power model for Abies pindrow were found to be better suited and valid for volume estimation in both forest ranges.ThesisItem Open Access Statistical Evaluation of Agricultural Development Pattern at District Level in Jammu and Kashmir(SKUAST Kashmir, 2021) Jadhav, Aaditya Pandurang; Nazir, Dr. NageenaAgricultural development is multidimensional and integrated approach which strengthened economic, social and cultural prosperity of state and country as well as ensures the quality improvement in standard of living. Study of agricultural development in Jammu and Kashmir serves a purpose because of several reasons. Since, Jammu and Kashmir have a great agricultural background, agriculture occupies an important place in the socio-economic prosperity of J & K and has great influenced on the growth rate of J & K economy. The prevailing environmental conditions, varying topography, farming practices, adoption of technology, irrigation facilities, farmer attitude etc. were the factors which create the marked regional difference in pace of agricultural development in different district of Jammu and Kashmir. Evaluation of agricultural development pattern at district level will help to identify potential source and factors responsible for development as well as the stand of each district in comparison with other. An attempt has been made in the present study to evaluate agricultural development at district level in Jammu and Kashmir constructing composite index based on 31 indicators. The method of principal component analysis was adopted. The value of indicators was standardised, assigned weights to the indicators using component factors loading and eigen values and composite index was formed for each of 22 districts of Jammu and Kashmir. Anantnag district has received highest index while as Ramban district has received the least index. The districts were classified into three categories namely, highly developed (> mean + ½ SD), moderately developed (mean – ½ SD) to (mean + ½ SD) and under developed (< mean – ½ SD) based on composite index. The highly developed group includes 8 districts namely Anantnag, Baramulla, Budgum Pulwama, Kupwara, Jammu, Kulgum, Kathua. Six districts namely, Udhampur, Shopian, Rajouri, Doda, Poonch and Bandipora fall in moderately developed category. The under developed category contain 8 districts namely, Ganderbal, Srinagar, Samba, Resai, Kargil, Leh, Kishtwar and Ramban. To identify the factors affecting agricultural development across districts multivariate analysis of variance (MANOVA) was performed. The results of MANOVA test concluded that the following indicators viz., percentage of net sown area to total reported area, area under rice, area under fruit and vegetables, area under apple, area under fresh fruits, area under dry fruit, production of apple, production of walnut, total production of fresh fruit and total production of dry fruit were the factors which affect the agricultural development across the districts in Jammu and Kashmir.
- «
- 1 (current)
- 2
- 3
- »