Generalization ability of Extreme Learning Machine using different Sample Selection Methods

The recent explosion of data has triggered the need of data reduction for completing the effective data mining task in the process of knowledge discovery in databases (KDD). The process of instance selection (IS) plays a significant role for data reduction by eliminating the redundant, noisy, unreliable and irrelevant instances, which, in-turn reduces the computational resources, and helps to increase the capabilities and generalization abilities of the learning models. . This manuscript expounds the concept and functionalities of seven different instance selection techniques (i.e., ENN, AllKNN, MENN, ENNTh, Mul- tiEdit, NCNEdit, and RNG), and also evaluates their effectiveness by using single layer feed-forward neural network (SLFN), which is trained with extreme learning machine (ELM). Unlike traditional neural network, ELM randomly chooses the weights and biases of hidden layer nodes and analytically determines the weights of output layer node. The generalization ability of ELM is analyzed by using both original and reduced datasets. Experiment results depict that ELM provides better generalization with these IS methods.

Analysis of Road Traffic Accidents in the Punjab by Using Panel Count Data Models

Each year more than three thousand people die and get serious injuries in traffic accidents. Count data model provide more precise tools for planners and decision makers to conduct proactive road safety planning.We analyzed the exploratory research of Road Traffic Accidents (RTAs) and furthermore explores the factors affecting the RTAs frequency in 36 districts of the Punjab over a time period of three years (July 1, 2013 June 30, 2016) with monthly data using panel count data models. Among the models considered, the random parameters Poisson panel count data model is found to fit the data best. The exploratory analysis shows that highly dense populated districts with large number of registered vehicles causes more accidents as compared to low density populated districts. It is found that, most of the variables used to control the variation in the frequency of RTAs counts play vital role with higher significance levels. The application of regression analysis and modeling of RTAs at district level in Punjab will help to identification of districts with high RTAs rates and this could help more efficient road safety management in the Punjab.

Some Improved Estimators of Population Mean Under Systematic Sampling

Auxiliary information is very important in constructing estimators for the population parameters for increasing the efficiency different sampling schemes. In this paper, we consider the problem of estimation of population mean using information on auxiliary variables in systematic sampling. We derive the expressions for the bias and mean squared error (MSE) of the suggested estimators up to the 1st degree of approximation. Proposed estimators are compared with usual mean estimator in systematic sampling scheme theoretically as well as empirically.

Finite Population Distribution Function Estimation Using Auxiliary Information Under Simple Random Sampling

In this paper, a new estimator for estimating the finite population distribution function(DF) are propose using supplementary information on the DF of the auxiliary variable under simple random sampling. A comparative study is conducted to compare, theoretically and numerically, the adapted distribution function estimators of Cochran (1940), Murthy (1967), Bahl and Tuteja (1991), Rao (1991), Singh et al. (2009) and Grover and Kaur (2014) with the proposed estimators. It is found that the proposed estimators always perform better than the adapted estimators in terms of MSE and percentage relative efficiency.

Statistical Modeling and Significance of Inputs Studies for the Rice Crop Productivity in Pakistan

Lower production of rice crop and increasing population trend is a tangible question for Pakistan. The current study based cross-sectional data collected from crop reporting service Punjab. The multiple linear regression (MLR) model is applied to investigate the significant factor for rice crop yield enhancement. The inputs variable i.e. owned land, seed rate, DAP, Urea, no. of water, no. of ploughs, no. of levels, crop life periods days, other fertilizers, variety super yes or no, seed type yes or no, spray no or yes and disease attack yes or no are studied in MLR model for rice productivity. All the factors found to be statistical significant except land, level, crop life period and super variety. Adj R2 is found to be 0.422 and it is good fit for cross-sectional data.

Socio-Economic Factors Affecting The Maternal Health In Rural Areas Of District Layyah, Pakistan

The present quantitative study was conducted to explore the socio-economic factors affecting on the maternal health in the rural areas of District Layyah. The universe of the present study consisted of the all the fertile women those were able to reproduce. 150 women were selected from the rural areas of the research areas through systematic random sampling. Description of the data and analysis was done through SPSS. It was concluded that the early marriages, low level of education and income, unavailability of the maternal homes and general hospitals, far away of the hospitals, and the absence of doctors and gynecologists in the rural areas and the traditional methods of delivery cases are the major causes of the low level of maternal health in the rural areas of the study area. It was recommended that to improve the maternal health education and income level should be raised and awareness should be given in the study areas.

Using Auxiliary Information More Efficiently in Population Variance Estimation - A New Family of Estimators

In this article, we have suggested estimation of variance in finite population by using known values of parameter related to auxiliary information such as rank and second raw moment of auxiliary variable in stratified random sampling. The expression for the bias and mean squared error (MSE) of the suggested estimator are obtained up to first order of approximation. The proposed estimator is efficient comparatively various other estimators. A numerical and theoretical study are performed to support the suggested estimator.

Major Socio-Economic Factors Responsible for Crimes

Crime in Pakistan has been increasing day by day with the passage of time. This research discussed two particularly influential approaches to the explanation of the role of Socio-Economic factors that are responsible for crimes. One of these approaches emphasizes social orientations and the other economic deprivation. Certain groups allegedly endorse values that are supportive of criminal behavior. In this study Quantitative research method is used and universe of this study is district jail Sargodha. Data is collected through questionnaire and 110 respondents included in this study which is selected after applying simple random sampling technique and through SPSS (statistical packages for social sciences) data analysis is comprises. Even though statistical significance is often difficult and problematic but this technique is used to logically and scientifically approve the factors including social and economic which create a sense of disturbance and responsible for creating the rebellion traits in that individual. These result shows crimes have directed attention to poverty, injustice, unemployment, lawlessness, anomic situation, general economic inequality, and inequality. So socio-economic factors often present and held responsible behind every criminal act that is done by a human. Now as the result the people chose the path of crime that satisfied and pulled out him from the tense situation temporarily but they have to pay the consequences of their criminal actions in the form of a few year prison in corresponding to a small type of crime and sometimes its declared as a lifetime time prison and death penalty according to the severity and intensity of crime.

Regional Rainfall Frequency Analysis on Seven Sites of Punjab, Pakistan using L-moments

The hydro-meteorological variables of extreme rainfall are not easy to explain due to unexpected changes in climate and varied usage of water with growing population. Regional rainfall frequency analysis is the one such method that is useful for the requirement of more accurate estimates of rainfall yearly or desineally for the regions having lack of fresh water resources. The series of Annual Maximum Monthly Rainfall Totals (AMMRT) has been used for the seven sites of northern Punjab, Pakistan using L-moments. The results of different test, the run test, lag-1 correlation and Mann-Whitney U test illustrate that the data series of the seven sites of northern Punjab were found random and independently and identically distributed and have no serial correlation. Heterogeneity measure exposed that the region is homogeneous and discordancy measure gives the evidence that no site is discordant among the seven. The result of goodness of fit test including L-moment Ratio diagrams, ZDIST statistic and Mean Absolute Deviation Index exposed the Pearson Type III (PE3), Generalized Normal (GNO) and Generalized Extreme Value(GEV) are best suitable of the regional distribution for the quantiles estimation. The quantiles estimates obtained for different return periods. A linear regression model was developed with good fit between the at site characteristics and the mean of the AMMRT of the sites. The estimates of the study may be used for the estimation of the rainfall quantiles of the seven sites for different return periods. The estimates will be useful to design future preventive measures for the harmful impact of hydro meteorological events at these sites in Punjab Pakistan.

Socoi-Economic and Cultural Constraints in Higher Education of Females in Pakistan: (A Case Study of Dera Ghazi Khan)

Education is the requirement of all men and women to compete in this era of knowledge as without education, survival is very difficult for both male and female. In rural areas of Pakistan female being subjugated under the patriarchal eastern society are most vulnerable condition. Their vulnerability is at worst in the areas where mainly tribal life form exists. Dera Ghazi Khan is one of the districts of Punjab, where tribal life style exists. Women under tribal regulations have a little access to education while very low access to higher education. The study was conducted in Dera Ghazi Khan, Punjab, Pakistan and it was conducted by adopting quantitative research method; a structured questionnaire was developed to collect data from 400 female. The study was an effort to explore the issues which restrict female from higher education. The major findings revealed that female face many problems and constraints in their higher education, such as long distance, less awareness about higher education, domestic responsibilities, early marriage and also female face issues regarding their security.