Journal Volume: 75      No.: 1     Year: 2020
S.No Title Abstract Download
1 Detection of Multiple Outliers in Time Series An Application to Rice Yield Data
Author: Gopal Saha, Ranjit Kumar Paul and L.M. Bhar      Pages: 1-7
Detection of outliers in time series data is a key component of data analysis. As the presence of outlier have a serious effect on model identification statistics, therefore conclusions drawn through analyzing the data series contaminated with outliers may be erroneous. It is, therefore, important to identify the time points where outliers are present and then remove the effect of the outliers from the corresponding series. The present paper considers the detection of outliers in time series data. An iterative method based on the procedure proposed by Chang and Tiao (1983) along with use of robust estimate of error variance is discussed. The power of this iterative procedure in detecting outliers is also investigated. The methodology is illustrated using rice yield data for all India during 1950-2013. The result of the study clearly indicates outlier detection technique using the robust estimate of error variance can successfully detect all the outliers present in the data series. Keywords: Autoregressive moving average model, Intervention model, Iterative estimation, Outlier.
2 Fitting of Size Biased Generalized Negative Binomial and Poisson Distributions on Crop Pests Data
Author: Chetan Kumar Saini, H.L. Sharma and Pushpendra Patel      Pages: 9-12
This paper is concerned with the fitting of size biased generalized negative binomial (SBGNBD) and size biased generalized Poisson distributions (SBGPD) on crop pest?s data. In these distributions, the recorded observations cannot be considered as a random sample from the original distributions due to the observations fall in the non-experimental, non-random and non-replicated categories where the truncated distributions which lie above or below a given threshold level of the distributions. The parameters involved in SBGNBD and SBGPD have been estimated by method of proportion of one?th cell (MPOC) and method of moments (MM). The distributions describe the data satisfactorily well. Keywords: Method of moments, MPOC, Parameter, ?2
3 Augmented Simplex-centroid Designs for Mixture Experiments
Author: Debopam Rakshit, B.N. Mandal, Rajender Parsad and Sukanta Dash      Pages: 19-23
For an experiment with mixtures, it is presumed that the response is dependent on the relative proportions of the components present in the mixture and it is invariant of the total amount of the mixture used in the experiment. More number of the design points from the interior of the simplex space can help for better exploration of the entire simplex space for the standard simplex-centroid designs. In this article, a new method for obtaining augmented points from the interior of the simplex space for three-component simplex-centroid designs has been proposed. A special property of equilateral triangles has been used for this purpose. D-efficiency and G-efficiency of these augmented designs are evaluated to measure the efficiency of these obtained designs. Keywords: Mixture experiments, Simplex-centroid design, Augmented, Componen
4 A Series of Factorial Row-Column Designs with Incomplete Rows and Columns
Author: Cini Varghese, Seema Jaggi, Kallol Sarkar and Mohd. Harun      Pages: 13-18
5 Measuring the Infrastructural Adequacy for Agriculture: A Comperative Analysis of Indian States
Author: Rajni Jain, Prem Chand, Priyanka Agarwal, Sulakshana Rao and Suresh Pal      Pages: 25-35
Agricultural infrastructure influences the yield as well as facilitates procurement, processing, preservation, and trade. This study developed a methodology for quantifying the status of both physical and institutional infrastructures for agriculture in India. Further, the study identified the relative state-level agricultural infrastructural adequacy status in the country based on secondary datasets. Spatial variation was observed in irrigation, markets, road, extension, credit, and storage infrastructure. Composite infrastructural classes highlighted that all the states in the country have infrastructural inadequacy in one or more parameters. Thus, improvement in agricultural infrastructure in the country calls for huge investments to enhance the income of the farmers. Keywords: Infrastructure adequacy, Agricultural sustainability, Enhancing farm income, Socio-economic suitability
6 Sampling Methodology for Estimation of Harvest and Post-harvest Losses of Major Crops and Commodities
Author: Tauqueer Ahmad, Anil Rai, Prachi Misra Sahoo, S.N. Jha, R.K. Vishwakarma      Pages: 37-46
In developing countries, efficient use of food materials produced and saving them as much as possible is one of the ways of achieving the target of ensuring availability of food to the masses. In order to make strategies for reducing the losses, knowledge of extent of losses and their reasons is essential. Based on the literature reviewed on estimation of harvest and post-harvest losses, it has been observed that most of the studies did not follow standard statistical methods and thus may not reflect the accurate scenario of extent of losses at national level. Therefore, an appropriate sampling methodology for estimation of quantitative harvest and post-harvest losses of major crops and commodities has been developed. Estimates of percentage loss along with percentage standard error of the estimates at agro-climatic zone level and national level have been obtained using this developed methodology. This methodology provides reliable estimates of quantitative harvest and post-harvest losses of 45 crops and commodities in India through an integrated national level survey conducted during 2012-2014. The developed methodology generates reliable estimates of losses in various operations and storages in different channels. Keywords: Harvest loss, Post-harvest loss, Sampling design, Estimation procedure, Survey, Sampling methodology, Operations, Channels, Percentage
7 Calibration Estimator of Finite Population Mean using Auxiliary Information under Adaptive Cluster Sampling
Author: Ankur Biswas, Raju Kumar, Deepak Singh and Pradip Basak      Pages: 47-53
Adaptive cluster sampling (ACS) technique is usually used for estimation of the abundance of an exclusive, clustered biological population. Commonly, neighbouring units are added to the sample if it satisfies a pre-determined criterion. Use of auxiliary information to increase the precision of estimators is a very general practice. This paper deals with the use of auxiliary information for the development of efficient estimator of finite population mean under ACS design using the well-known Calibration Approach given by Deville and Särndal (1992). The statistical performance of the calibration estimators of population mean under ACS are evaluated through a simulation study with respect to conventional Horvitz Thomson (HT) estimator of population mean which do not utilize the auxiliary information. The results of the simulation study conducted on a rare and clustered population often cited in Smith et al. (1995) show that proposed calibration estimators are more efficient than conventional HT estimator of the population mean under ACS with respect to percentage Relative Bias (%RB) and percentage Relative Root Mean Squared Error (%RRMSE). Keywords: Calibration, Auxiliary information, Rare attribute, Clustered.
8 Measuring Price Transmission, Causality and Impulse Response: An Empirical Evidence from Major Potato Markets in India
Author: Soumik Dey, Kanchan Sinha , Arnab Kumar Chand , Pramit Pandit , Herojit Singh and P.K. Sahu      Pages: 55-62
Commodity price transmission plays a pivotal role in determining the market leader among the comparable markets and for identification of the direction of price movements. This study makes an attempt to examine the cointegration and price transmission mechanism among three wholesale potato markets in India viz., Mumbai, Agra and Burdwan, during the period from January, 2016 to December, 2018 collected from www.agmarketnet. Several econometric tools viz. stationarity tests (ADF and PP test), Johansen?s cointegration test, Granger?s causality test, Vector Error Correction Model (VECM), Impulse response function etc. are utilized to fulfil the objectives of the study. Johansen?s cointegration test establishes the presence of long run equilibrium relationship among the markets with a single cointegrating equation. The study confirmed no price transmission occurs from Burdwan to Mumbai as there exist only unidirectional causality from Mumbai to Burdwan. From the VECM, it has been observed that the speed of adjustment (error correction term) from deviation towards the long run equilibrium for Agra and Mumbai markets are significant with daily price adjustment of 4.14% and 7.10% respectively, whereas, Burdwan market fails to establish any sort of significant behaviour. The impulse response curve also fails to establish any long-run association from Mumbai and Agra market towards Burdwan market. Based on all the market responses, both Agra and Mumbai market can be realized as the price leader as they influence the prices of all other markets. Keywords: Price transmission, Cointegration, Impulse response function, VECM.
9 Hindi Supplement vol 75 (01)
Author: ISAS      Pages: 5
10 Construction of Two Level Balanced and Nearly Balanced Optimal Supersaturated Designs
Author: Parvez Mallick, Jit Shankar Basak, A. Dutta, H. Das and A. Majumder      Pages: 63-73
Supersaturated designs (SSDs) are very useful for screening experiments with many factors using only a few runs or design points. The widely accepted criteria for optimality of two level SSD is the E(s2) measure, where the design matrix Xd has the restriction that either each column sum will be zero for balanced supersaturated designs or each column sum will be ±1 for nearly balanced designs (Gupta, 2010). Several researchers have constructed many two level balanced and nearly balanced SSDs for different combinations of m and n (m stands for number of factors and n number of runs; m ? n). The solutions of almost all the available balanced and nearly balanced SSDs are presented in ?Design Resource Server? of IASRI website. Some new methods are developed for construction of new balanced and nearly balanced SSDs. In the first part of the article, some new methods of construction of balanced and nearly balanced supersaturated designs have been presented. The methods yield some new balanced and nearly balanced optimum supersaturated designs which are not yet reported in the available literature. Many available supersaturated designs can also be constructed from these methods; in the sense these methods are more general. The developed designs are examined by sharper lower bounds of E(s2) measures (Suen and Das, 2010). The design points or solutions of some designs are given in Appendix I. In the second part, new methods for construction of two level balanced and nearly balanced supersaturated designs (master SSDs) involving maximum possible number (mmax) of factors for any particular number of runs (n), are presented. A series of new SSDs are constructed from those master SSDs after deleting the similar columns of available SSDs. Keywords: Hadamard matrix, Supersaturated designs and Lower bounds of supersaturated designs
11 Particle Swarm Optimization based Multi-objective Optimization for Crop Planning: A Case Study of Bundelkhand
Author: Shbana Begam, Rajni Jain, Alka Arora and Sudeep Marwaha      Pages: 75-85
Indian agriculture and its allied sectors are undeniably the major livelihood source in India, especially in rural areas, and it heavily depends on natural resources, climatic condition etc. The continuous depletion of natural resources and unpredictable climatic condition can cause low productivity growth and food security issues. A modified cropping pattern can be helpful to improve the net profit by utilizing minimum resources. A crop planning model is proposed here for the optimal allocation of resources under a water-deficit region like Bundelkhand. This study presents a Multi-objective Particle Swarm Optimization using Crowding Distance (MOPSOCD) which is an evolutionary algorithm to solve the constrained bi-objective crop planning problem. The objective functions of the model are maximizing total net returns and minimize the net water requirements. Maximum and minimum available land area, cropping area for various crops were considered as constraints of the model. The optimized results obtained from got a better crop planning strategy using the presented method. Overall, multi-objective optimization technique using PSO with crowding distance effectively improve optimal area allocation for water deficit region Bundelkhand with high range of net return and utilizes low water. Keywords: Crowding distance, Crop planning, Genetic algorithm, Multi-objective optimization, Particle swarm optimization.