Most of agricultural data are collected for a long period of time and need high attention. If one record has not been recorded, then the data will become incomplete. In Ethiopia, agricultural researchers have often been challenged by incomplete data. Different simulation techniques with different approximation capability have been used to solve this problem. As a result, this study is aimed to compare which computer based simulation techniques approximate the results of the previously accomplished researches of milk production traits. 15 years of data from Debre Zeit Research Station of the International Livestock Research Institute and Holetta Agricultural Research Centre of the Ethiopian Institute of Agricultural Research have been used for this study. We compared the two most familiar simulation techniques namely Monte Carlo and bootstrap simulations by using the results of linear mixed model fitted for each dataset. We found that both Monte Carlo and bootstrap simulations can approximate the farm and genetic group effects equally. Lactation length and daily milk yield are found to be significant (P < 0.0001) in both simulation techniques. Unlike for bootstrap simulation, season and period of calving are found to be significant for Monte Carlo simulation. On the basis of the findings, this study reached a conclusion that Monte Carlo simulation has a better approximation. Keywords: Monte Carlo, Bootstrap, Simulation, Model, Linear mixed model, Milk yield.

  Commodity future has a vital role to play in any economy as the future contracts perform two important functions of price discovery and price risk management. The present study has been undertaken to examine whether the future and cash markets follow the efficiency criterion in trading Soybean for discovering better price. Seven non-overlapping future contracts maturing on March 2008 to September 2010 and secured at NCDEX has been examined. Johansen?s cointegration test (1988) between future and spot price at Indore was carried out for each future contract of Soybean. The future and spot markets in NCDEX exchange are cointegrated and sharing a long run relationship. The two statistical tests, Trace Statistics and Eigen Value Statistics confirm the relationship of short and long run between spot and future price of soybean. There is a causality flow from future markets towards spot markets indicating information flow from future to spot markets. At the same time, there is also a reverse information flow happening in some contracts signifying price discovery in both future and spot markets. This finding, to a large extent, answers to the apprehensions of destabilizing impact of commodity future markets in India. The Johansen?s vector error correction model (VECM) indicates that the future market leads the spot market in most of the contracts whereas in two contracts spot prices also tends to discover new information more rapidly than future prices. Keywords: Cointegration, Price discovery, Risk management, Unit Root.

  Biological organism always tends to behave non-linearly contrary to linear growth as perceived in most of the data analysis procedure. In the present communication, a simple nonlinear logistic growth model has been developed to describe the population dynamics of incidence of downy mildew in grapes (cv. Anab-X-Sahai) so as to workout quantitative information about the biological parameters concerning intrinsic infection rate and maximum mildew severity over time-epoch. Statistical analysis of disease severity data over time period for three years (2004-05 to 2006-07) using non linear growth models revealed that 98% of the variability in disease progression over time-epoch was captured by nonlinear models. Nonlinear models developed were then used to construct area under disease progression over time period. Results showed that, in general, the rate of disease severity was maximum during fifth- sixth week after fore-pruning, calling for appropriate management strategies for controlling the disease within the period identified, thus avoiding crop loss. Before taking final conclusion about the model, the model- generated residuals were tested for their robustness using statistical techniques. SAS Programming codes were constructed to develop the nonlinear growth models. Keywords: Coefficient of determination, Downy mildew, Gompertz model, Grapes, Logistic model, SAS programming, Weather factors

  In surveys, sample size planning is important for achieving precise estimates at a low cost. However, this issue is not adequately addressed for group testing data obtained from a three-stage sampling process. In this study, we obtained the optimal allocation of localities (l), fields (m) and pools per field (g) in a three-stage group testing survey for a given pool size (s). These optimal values were obtained under the assumption of equal locality and field sizes. To handle the unequal sample size case, we derived the relative efficiency (RE) of unequal versus equal locality and field sizes to estimate the proportion. By multiplying the sample of localities and fields obtained assuming equal cluster size by the inverse of the corresponding REs, we adjusted the sample size required in the context of unequal localities and field sizes. We also show the adjustments needed for correctly allocating localities and fields in order to estimate the required budget and achieve a certain power or precision. In surveys, sample size planning is important for achieving precise estimates at a low cost. However, this issue is not adequately addressed for group testing data obtained from a three-stage sampling process. In this study, we obtained the optimal allocation of localities (l), fields (m) and pools per field (g) in a three-stage group testing survey for a given pool size (s). These optimal values were obtained under the assumption of equal locality and field sizes. To handle the unequal sample size case, we derived the relative efficiency (RE) of unequal versus equal locality and field sizes to estimate the proportion. By multiplying the sample of localities and fields obtained assuming equal cluster size by the inverse of the corresponding REs, we adjusted the sample size required in the context of unequal localities and field sizes. We also show the adjustments needed for correctly allocating localities and fields in order to estimate the required budget and achieve a certain power or precision.

  In this article we demonstrate an application of small area estimation technique to produce district level estimates of crop yield for three major crops of the State of Uttar Pradesh using the Improvement of Crop Statistics Scheme data and the auxiliary data from various secondary sources. In particular, we use a spatial model for small area estimation to improve the district level crop yield estimates. The results show improvement in the district level crop yield estimates due to use of spatial information in small area estimation. Keywords: Crop cutting experiments, Improvement of Crop Statistics, District level estimates, Small area estimation, Spatial model.

  Two statistics for detecting outliers in designed experiments with correlated errors have been developed. These statistics are Cook-statistic and AP-statistic. General expressions of these statistics for detecting any t outliers have been obtained. Equal correlation structure has been considered for general variance-covariance matrix. Developed Cook-statistic has been illustrated with an example. However, case of occurrence of a single outlier has been considered in the example. Keywords: Cook-statistic, AP-statistic, Outlier, Block design, Correlated error, Autocorrelation.

  The information related to crop area and crop production have been playing a vital role in planning and allocations of resources for the development of agriculture sector. In this paper, Artificial Neural Networks (ANN) has been used to predict the crop area and crop production for the crops Summer Rice and Maize of Upper Brahmaputra Valley (UBV) Zone of Assam as an alternative statistical tool. Multilayer Perceptron (MLP) with single hidden layer and Radial Basis Function (RBF) network have been trained with the secondary data of the crop area, crop production and meteorological data. The appropriate configuration for each of the network model is identified. The performance of the constructed ANN models has been measured using Root Mean Squared Errors (RMSE) and Correlation Coefficients (CC). The predictive accuracy of the developed ANN models has been compared with Multiple Linear Regression (MLR) Model. The performance comparison show that the constructed ANN-MLP and ANN-RBF models outperform MLR model. Sensitivity analysis has been performed for Prediction of Summer Rice and Maize production and results show that technology index is the most sensitive parameter followed by rainfall index for production of Summer Rice Production while temperature (maximum) is the most sensitive parameter for prediction of the crop Maize production followed by technology index for the UBV Zone of Assam with the considered secondary data. Keywords: ANN, MLP, RBF, Crop Area, Crop Production.

  The ?green revolution? era in India had largely by-passed the rainfed agriculture. In order to achieve overall development of agriculture, it is essential to bridge the yield gaps, enhance profitability, minimize risk and improve the livelihoods of millions of people dependent on rainfed agriculture. Therefore, regionally differentiated interventions befitting to the natural resource endowment are need of the hour to meet the current challenges. Further, the resources available for meeting such demands are naturally scarce. Therefore it is important to prioritize the rainfed areas and identify the possible interventions for formulating developmental programmes for correcting the plight in prioritized areas. Previous studies with this aim largely focused on variables like rainfall, irrigation etc. The present study made an attempt to prioritize the rainfed areas including variables like drought proneness and degraded and waste lands, water holding capacity of soil, ground water utilization status as well. For this purpose, district which is the lowest level administrative unit for which reasonable statistical data are available for planning and evaluation is considered as unit of study. A composite index, namely, Natural Resources Index (NRI) was developed by combining various facets of natural resource availability. Based on NRI score, the districts were ranked such that districts with low score (weak resource base) get high priority for resource conservation and upliftment. Status of natural resources (NRI) is relatively low in western, central and southern part of the country extending from Haryana to Tamil Nadu with exception of west coast region of Karnataka and Kerala. On the other hand, the NRI is relatively high in eastern parts of India particularly in West Bengal, Bihar and Orissa. The information on various facets as summarised in the form of indicators serves as guiding tool for identifying appropriate interventions. Keywords: Rainfed areas, Prioritization, District, Natural resources.

  Compound growth rate is widely employed in the field of Agriculture as it has important policy implications. Generally, this is computed by assuming that the path of response variable can be described by monotonically non-decreasing nonlinear growth models, like Malthus model and logistic model. However, in reality, data sets in agriculture need not always depict steady upward or downward movement over time, i.e. sometimes these are non-monotonic in nature. In such cases, it is not appropriate to employ above growth models. In this article, three nonlinear growth models, viz. over-damped, under-damped and critically-damped are considered, which have the capability to describe increasing and then decreasing or vice-versa type of behaviour of the response variable. The methodology to estimate compound growth rate by using these growth models is discussed. As it is very difficult to apply, an online user-friendly web-based application, viz. WebECGR package is developed. Finally, as an illustration, this package is employed for estimation of compound growth rate for India?s total lentil production data during the period 1980-81 to 2010-11. Keywords: Compound growth rate, Critically damped model, India?s total lentil production, Non-monotonic situations, WebECGR package.

  Artificial Neural Networks (ANNs) are non-linear structures used for prediction and classification problems. ANNs identify and learn correlated patterns between input data sets and corresponding target values. Trained ANNs are used to predict the outcomes of independent variables. Over fitting and under fitting are two major problems that may arise in ANNs. When two or more predictor variables in a model are highly correlated, called as multi-collinearity, they provide redundant information about the response and leads to overtraining. This problem is handled by using ANN with weight decay algorithm. Many software are available for analyzing the data using ANN but either they are very expensive or difficult to use. This study describes a web based software for back propagation neural networks with weight decay algorithm. This software is useful for statisticians and researchers implementing ANNs for various data mining task and facing non-convergence problem. Keywords: Artificial neural networks, Software, Web, Weight decay.

   N.A

   N.A

   N.A

   N.A