N.A

  Dr. Daroga Singh

  N.A

  The paper highlights some important contributions by the author jointly with others in development of some newer sampling designs which had ?intuition? as its basis and theory derived thereafter.

  While dealing with the observed or measured data in surveys or experiments, it is not uncommon to deal with vague, incomplete, and impreciseinformation for whatever reasons. In this regard, researchers have proposed various emerging approaches such as fuzzy, intuitionistic fuzzy andneutrosophic logic and analysis, which provide better understanding, analysis and interpretations of the data. Neutrosophic logic is an extension offuzzy logic where a variable x is described by triplet values, i.e., x = (t, i, f ) , where t is the degree of ?truth?, f is the degree of ?false? and i isthe level of ?indeterminacy? AlAita, Abdulrahmanand Aslam, Muhammad (2023). A neutrosophic data x can be expressed as 𝑥 = 𝑑 𝑖, where 𝑑 isthe determinate (sure) part of 𝑥, and 𝑖 is the indeterminate (unsure) part of 𝑥. Experimental design and analysis is a systematic, rigorous approach toproblem solving that applies principles and techniques at the data collection stage so as to ensure the generation of valid, defensible, and supportableconclusions. Latin square designs (LSDs) are used to compare treatment factor levels represented by the Latin letters and using two blocking factorsin rows and columns to simultaneously control two sources of nuisance variability. In this paper, we will define a neutrosophic Latin square design(NLSD), neutrosophic LSD model and consider the neutrosophic analysis of the NLSD experimental data for testing the abrasion resistance of rubber coatedfabric in a Martindale wear tester.

  In the present paper, some important contributions on estimation of finite population parameters under model-based prediction approach are reviewed.Following Scott et al. (1978), a BLU predictor of population total in model-based prediction approach under the model (0,1: 2 ) hk ξ x in stratifiedsampling when the slope of the model is common across the strata is constructed. Its robustness and optimality is studied when some generalpolynomial model of degree j, i.e., 20 1 ( , ,....., : ) J hk ξ δ δ δ x is true in real practice. It has been found that the proposed predictor is robust and optimalfor stratified balanced sample and it is also more efficient than the predictor developed by Scott et al. (1978) when slopes are uncommon across thestratum.

  For a Direct Survey on innocuous characteristics Chebyshev?s inequality is helpful in prescribing the size of a sample in a survey. An extension of the same to cover stigmatizing features in Randomized Response (RR) survey is not smooth enough. Different situations are illustrated and solutions proposed.

  We are interested in unbiased estimation of the unknown size of a finite population. The population units [also called Ultimate Units (UUs)] are not directly accessible in any way. We can only have access to any UU via an appropriate Reference Unit [RU] only if it captures the UU in question. The literature is scanty and the state of knowledge also seems to be imperfect. We aim at providing an overview of the literature in this fascinating area of research and its application.

  Linear integer programming is a widely used optimization technique to solve various real life problems. The purpose of this article is to present some innovative applications of linear integer programming in the area of design of experiments and sample surveys. It is demonstrated how construction problems of various block designs and different classes of sampling plans can be solved using linear integer programming formulations.

  The coefficient of variation (CV) has been used in different disciplines with varied purpose related to variation in quantitative measurements. Statistical properties of CV have been studied by various researchers. Recently, a paper by the authors featuring an investigation of the posterior distribution of a common CV for inverse Gaussian populations with priors obtained through some empirical fitting procedure was presented by YPC at the 2022 ISBA World Meeting, June 26-July 1, 2022, held in Montreal, Canada. Some of these results along with other current developments by the authors on the topic were also reviewed during an invited presentation by MS in the honor of Dr. Daroga Singh at the 73rd Annual Conference of ISAS Conference held at the Sher-e-Kashmir University of Agricultural Sciences and Technology of Kashmir, November 14-16, 2022. The purpose of this paper is to summarize and discuss three key papers on estimation and testing of CV from inverse Gaussian distribution focussed on ? a common CV (from multiple populations under frequentist framework), CV for a single population with Bayesian framework and a common CV from multiple populations and Bayesian framework, reviewed at this conference.

  I consider it a great honor to be called upon to prepare a paper on birth centenary of Dr. Daroga Singh, Former Director of the Institute. I am grateful to ISAS and especially Dr. Padam Singh, Executive President, ISAS for giving me this opportunity. I am greatly indebted to Dr. Daroga Singh for his generosity, kindness, guidance and support right from my student days at IASRI and his inspiration and guidance as Chairman of my Advisory Committee for both my M.Sc. Degree and Ph.D Degree at IASRI and in Delhi University. After I joined service at IASRI, Dr. Daroga Singh remained a great inspiration and support for which I will remain indebted to him throughout my life. During 1980?s, NATP Programme sponsored by UNDP was implemented at ICAR and its research institutes to give a thrust to application of the new and advanced technologies in different fields of Agricultural research to boost the production and research. During this period, Dr. Prem Narain has been the Director of the Institute. He had a great vision and zest to work and during his time, the research activities of the Institute expanded vastly. He had great belief in the enormous strength of the Remote Sensing technology and hence it was recognized as a new thrust area of research under NATP and I was given the responsibility to initiate research programs in this field. As Dr. D. Singh had been great supporter of new technology therefore, we have therefore decided to present work done at IASRI in the area of Remote Sensing and Geo-Informatics in this paper, which is dedicated to him.

  This article presents the state of art review in the domain of weather based pre-harvest crop yield forecasting models. To begin with, it discusses the various approaches evolved over time for construction of weather indices from weather variables. Owning to complex relationship between crop yield and weather variables, various models have been proposed in literature. It has been documented starting with widely accepted weather based regression model followed by its improvements in variable selection and relaxation of assumptions. LASSO, ANN and random forest have been proposed as an improved method for variable selection, whereas Bayesian framework provides a way out when the data fails to satisfy the usual regression assumptions. The other developments such as use of complex polynomial through GMDH, principal component regression, discriminant function analysis and water balance technique have also been discussed. Further, we have briefly documented a webtool named Weather Indices based Automated Yield Forecasting System (WIAYFS) which has been developed for ease of implementation and reaching out to more researchers and users.

  This study proposes a novel approach for multi-step time series forecasting using a stacked long-short term memory (LSTM) sequence-to-sequence autoencoder (LSTM-SAE) to handle the volatility of edible oil prices in the Indian market. The approach was implemented on Ruchi Soya Ltd. stock price dataset and compared with other deep learning models like Gated Recurrent Unit (GRU), LSTM, and Bi-directional LSTM. The LSTM- SAE outperformed other models in closing price prediction based on evaluation metrics like Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The proposed approach has significant implications for stakeholders in the edible oil and oilseeds industry, including farmers, traders, and policymakers.

  This paper aims to develop a hybrid model using ensemble empirical mode decomposition (EEMD) as a decomposition technique and time-delay neural network (TDNN) as a forecasting technique to predict non-stationary and nonlinear agricultural price series. The EEMD first decomposes the agricultural price series into several intrinsic mode functions (IMFs) and a single residual. Further, the resulting IMFs and residual series are grouped into high frequency, low frequency, and a trend component with similar frequency characteristics to capture numerous coexisting hidden factors using the fine-to-coarse reconstruction (FCR) algorithm. After that, a TDNN with a single hidden layer is built to separately forecast each of the three nonlinear components. Finally, the prediction results of all three components are summed up to obtain a final output as the forecast of the original price series. The performance of the proposed hybrid EEMD-FCR-TDNN model is empirically evaluated by comparing it with several benchmark models, including the TDNN model and decomposition-ensemble hybrid models without reconstruction using monthly international maize and soybean oil price series. The results validate that the EEMD-FCR-TDNN model can significantly outperform the other models in terms of both level and directional prediction accuracy with lower computational cost.

  Kaur et al. (2000) suggested an optimal allocation model for symmetric distribution in ranked set sampling (RSS). This model is not much applicable due to its dependency upon extreme or mid order statistic only. In this paper, an attempt to make a ?favourable? and near optimal allocation has been made by allocating each rank order at least once and considering the opposite behavior of Neyman allocation for symmetric distribution. The case of perfect ranking is considered. The utility of the proposed model in terms of relative precision has been shown for some symmetric distributions. Favourable model outperforms both for equal and Neyman?s allocations and quite close to the optimal model for each set size. The model will provide a practical approach in situations where RSS is likely to lead to an improvement over simple random sampling and the underlying distribution is symmetric.

  In India, Food and Agriculture Organization of the United Nations (FAO) was implementing a project ?Strengthening Agriculture Market Information System (AMIS) in India using Innovative Methods and Digital Technology? and supporting the efforts of the Ministry of Agriculture and Farmers Welfare, Govt. of India. This project identified the potential of improving the data coverage on ?on-farm? post-harvest management of food grains through Input Survey carried out in Agriculture Census. Therefore, a pilot study on private food grains stock estimation at farm level aligned with Input Survey of Agriculture Census in India funded by FAO-India was conducted by ICAR-Indian Agricultural Statistics Research Institute (ICAR‑IASRI). Under this study, a suitable sampling methodology aligned with existing Input Survey for estimation of private food grain stock at farm level has been developed. A suitable questionnaire aligned with existing Input Survey of Agriculture Census has been developed covering different food grains stock at farm level. Under this study, a pilot survey was conducted in two states namely Haryana and Madhya Pradesh. The four crops under AMIS study i.e. wheat, paddy, maize and soybean along with pulses were covered under this pilot survey. The data was collected for all the three seasons. The estimates of food grains stock, pre-harvest opening stock, production obtained, quantity sold, quantity stored, quantity disposed and percentage stock at farm level were obtained along with its percentage Coefficient of Variation (% CV) and were found to be reasonably good for overall size classes. Therefore, it is expected that for overall holding size classes, the proposed methodology will provide farm level reliable estimates of food grains stock at district level. The study has established the feasibility of inclusion of developed questionnaire in the future Input Survey of Agriculture Census in India in order to estimate the food grains stock at farm level which will bridge the gap on private food grains stock in on-farm and off-farm domains of the supply chain.

  In this paper, mean estimators under ranked set sampling are reviewed. In this paper, we have also presented some improved novel classes of estimators for estimating the population mean using auxiliary variable under ranked set sampling. We have derived the expressions for bias and mean squared error of the proposed estimators up to the first order of approximation and the proposed classes of estimators are found to be more efficient than the other estimators in this study. In an attempt to verify the efficiencies of proposed estimators, theoretical results are supported by empirical study.

  Coefficient of variation (CV) is used to measure the heterogeneity of the experimental field in the context of agricultural research including sugarcane experiments.Smaller the CV value of cane yield, higher is the reliability of the experiment. In all the zones of All India Coordinated Research Project on Sugarcane - AICRP(S), upper limit of CV for cane yield is higher for early varietal trials in comparison tothe mid to late varietal trials of sugarcane in India. Similar trend was also found for upper limit of CV for sucrose (%). Upper limit of CV (%) for cane yield was found highest in Peninsular zone as around 10% in early varietal trial followed by North central Zone. It is also observed as lowest in North West Zone, which is around 8.5 % both for early and mid-late varietal trials. Same trend was also observed for sucrose, which around 3% at 1%, 5% and 10% level of significance. Overall for all the trial of cane yield, it is estimated as 9% for cane yield in the country at all the three level of significance. For plant crop, it is estimated as 9% at all level of significance. For ratoon crop, it is estimated as 10% at all level of significance.

  Double sampling for stratification is a statistical technique used in survey sampling to improve the efficiency of estimating population parameters estimates. In the present paper we have proposed a generalized dual to ratio-cum-product type estimator of population mean. The expression for the MSE (Mean Square Error) of the proposed estimator has been derived upto first degree of approximation. The comparison of the proposed estimator with Ige and Tripathi (1987) and Lone et al. (2020) estimators in double sampling for stratification indicated that proposed estimator is more efficient than the other estimators given in the literature. Motivated by Tailor and Lone (2014), the generalised version of the proposed estimator is also suggested in the present paper. The empirical study indicated that the proposed estimator is more efficient than the other estimators.

  This research work addresses bivariate auxiliary information-based logarithmic type direct and synthetic estimators for domain means in simple random sampling (SRS). The mean square error (MSE) of the suggested estimators is obtained, approximately to the first order. The efficiency standards by which the superiority of the suggested estimators is asserted are established. To demonstrate the superiority of the suggested estimators, a simulation investigation employing an artificially constructed normal population through the R programming language is also conducted. The analysis of real data from Swedish municipalities and the paddy crop acreage in the Mohanlal Ganj tehsil, Uttar Pradesh, India, also provides some applicability for the suggested estimators.

  Not Applicable