-
Divide China's Economic Regions in 2019 Based on Cluster Analysis and Principal Component Analysis
Zhichao Zhan,
Yongquan Jin,
Meihua Dong
Issue:
Volume 7, Issue 4, December 2021
Pages:
83-88
Received:
5 July 2021
Accepted:
26 July 2021
Published:
5 November 2021
Abstract: In recent years, China's economic development has been very rapid. While China is developing rapidly, each province has contributed its share, but in different regions, economic development is different. Different regions must have advantages in different aspects, so in order to divide China's 31 provinces into different categories. In order to get the ranking of the provinces that have the greatest impact on China's economy. We first adopt the method of principal component analysis to reduce the dimensions of 11 variables that affect the economic factors of each province, and obtain two principal components to reflect all sample information. Then, perform dimensionality reduction and cluster analysis on the obtained data, and use the sum of squared variance (WARD) method to perform cluster analysis on the two principal components. Finally, the social development of 31 provinces in my country is divided into 4 categories. It is concluded that Beijing and Shanghai are first-level developed provinces, Jiangsu and Guangdong are second-level developed provinces, Hebei, Sichuan, Hunan, Shandong, Henan, Shanxi, and Hubei are third-level developed provinces, Tianjin, Hainan, Tibet, Qinghai, Ningxia, Inner Mongolia, Jilin, Gansu, Xinjiang, Fujian, Chongqing, Liaoning, Anhui, Shaanxi, Jiangxi, Guizhou, Yunnan, Heilongjiang, and Guangxi are four-tier developed provinces. I hope our results can help relevant departments.
Abstract: In recent years, China's economic development has been very rapid. While China is developing rapidly, each province has contributed its share, but in different regions, economic development is different. Different regions must have advantages in different aspects, so in order to divide China's 31 provinces into different categories. In order to get...
Show More
-
Predictors for Risk Factors of Diabetes: Binary Logistic Regression Model Approach
Usman Aliyu,
Abubakar Umar Bashar,
Umar Usman
Issue:
Volume 7, Issue 4, December 2021
Pages:
89-94
Received:
15 August 2021
Accepted:
22 September 2021
Published:
10 November 2021
Abstract: Although Diabetes is a metabolic disease that cause high blood sugar to mostly people of age 45 years and above whose body either doesn’t make enough insulin or can’t effectively use the insulin it does make. This research use Binary Logistic regression model to analyze predictors for risk factors of Diabetes in Kebbi state as well as build a suitable Binary Logistic regression model capable of finding the relationship among predictors of Diabetes and check which of the predictors are more suitable in predicting of Diabetes. The data used in this study were obtained as secondary data from Federal Medical Centre Birnin Kebbi and Sir Yahaya Memorial Hospital Birnin Kebbi, the data consist 500 people diagnosed for Diabetes out of which some happens to be diabetic positive while others are negative. The analysis was performed using statistical package for social sciences (SPSS Version 21) and it was discovered that; Ages, History of Diabetes, History of HBP/Hypertension, and Overweight/Obese are risk factors of Diabetes and Sex, Frequent urination/increase thirst and Fatigue or Muscle Pain were not risk factors of Diabetes as the model obtained was Logit (P(y=1))=-6.154 + 2.609Age + 2.457History of HBP/Hypertension + 1.307History of Diabetes + 1.237Overweight/Obese.
Abstract: Although Diabetes is a metabolic disease that cause high blood sugar to mostly people of age 45 years and above whose body either doesn’t make enough insulin or can’t effectively use the insulin it does make. This research use Binary Logistic regression model to analyze predictors for risk factors of Diabetes in Kebbi state as well as build a suita...
Show More
-
An Analysis of Corona Virus Disease (COVID-19) Predictors: Logistic Regression Model Approach
Usman Aliyu,
Abubakar Umar Bashar,
Umar Usman
Issue:
Volume 7, Issue 4, December 2021
Pages:
95-101
Received:
15 August 2021
Accepted:
17 September 2021
Published:
19 November 2021
Abstract: Although Corona Virus disease (COVID-19) is a contagious disease cause by severe acute respiratory syndrome which affects mostly people whose immune system are weak or not resistance to the disease, there exists no vaccine that is 100% effective for its cure though efforts are being intensify by researchers in discovering the vaccine as well as model for prediction of Corona Virus Disease. In this era of advanced information and communication technology, as well as evidence-based medicine, statistical modeling has become as necessary the medical practitioners who are interested in lasting solution to diagnosed problems. In this work a logistic regressions model has been proposed to serve the purpose. The data was obtained from Nigeria Centre for Disease Control (NCDC) and was analyzed using binary logistic regression model in which Corona Virus disease was considered as categorical dependant variable (COVID-19 status: chance of being positive or negative) and the predictors considered are; Age, any of either Headache or Vomiting, Fever, Sore throat/runny nose, Any of Cold, cough or sweating, Loss of Smell or taste, and Breathing Difficulties. The results shows the significant predictors for predicting Corona Virus Diseases are; Loss of Smell or taste, Breathing Difficulties, Fever, Sore throat or runny nose, Age, any of either Headache or Vomiting, and Any of Cold, cough or sweating. The logit model obtained was: Logit (P(y=1)) = -3.748 + 0.356 Age +2.938 any of either Headache or Vomiting + 0.752 Fever + 2.792 Sore throat or runny nose – 0.028 Any of Cold, cough or sweating + 1.872 Loss of Smell or taste + 0.844 Breathing Difficulties. So also from the same results, it was found among predictors that; Sex/Gender, Temperature >37.5 degree and Fatigue or Muscle Pain were not good predictors of Corona Virus disease.
Abstract: Although Corona Virus disease (COVID-19) is a contagious disease cause by severe acute respiratory syndrome which affects mostly people whose immune system are weak or not resistance to the disease, there exists no vaccine that is 100% effective for its cure though efforts are being intensify by researchers in discovering the vaccine as well as mod...
Show More
-
On Some Models for Wind Power Assessment in Yola, Nigeria
Gongsin Isaac Esbond,
Funmilayo Westnand Oshogboye Saporu
Issue:
Volume 7, Issue 4, December 2021
Pages:
102-107
Received:
25 September 2021
Accepted:
21 October 2021
Published:
23 November 2021
Abstract: Probability distributions are used in the evaluation of wind energy potentials to describe the wind speed characteristics of the chosen location for wind farm establishment. However, the Weibull distribution that is the most chosen by wind energy modelers may likely fail to properly describe the wind speed data of certain locations, or it may not be the best model to describe wind speed when compared to the fitness of other probability distributions. Thus, in this study, four probability distributions are fitted to wind speed data from Yola, Nigeria. They are the Weibull, the exponentiated Weibull, the generalized power Weibull and the exponentiated epsilon distributions; and, all provided good fit to the wind speed dataset. The exponentiated epsilon distribution is new and provided the best fit. These models are compared based on the relative likelihood gain per data point; it is found that there is about 5% gain by the other three probability distributions over the Weibull distribution. Hence all the three distributions can also be used as wind models. The estimated average wind speeds computed using the four models at various hub heights show that wind is sufficiently available to support a wind turbine with a cut-in speed of 3 m/s at hub heights 90 m above ground level. For the exponentiated-epsilon model, average wind speed of 3.68 m/s at hub height of 120 m above ground level can generate 6.11 W/m2 of electricity; and for a wind turbine of rotor diameter of 128 m with 12,868 m2 swept area, this amounts to 78.6 kW of electricity supply for a small-scale wind power project. Consequently, Yola holds a good potential for the establishment of a wind farm.
Abstract: Probability distributions are used in the evaluation of wind energy potentials to describe the wind speed characteristics of the chosen location for wind farm establishment. However, the Weibull distribution that is the most chosen by wind energy modelers may likely fail to properly describe the wind speed data of certain locations, or it may not b...
Show More
-
Likelihood-Based Confidence Intervals for the Parameters of a Simple Linear Regression Model with Cauchy Errors
Issue:
Volume 7, Issue 4, December 2021
Pages:
108-114
Received:
21 October 2021
Accepted:
12 November 2021
Published:
24 November 2021
Abstract: The estimators of the slope and the intercept of simple linear regression model with normal errors are normally distributed and their exact confidence intervals are constructed using the t-distribution. However, when the normality assumption is not fulfilled, it is not possible to obtain exact confidence intervals. The Wald method of interval estimation is commonly used to provide approximate confidence intervals in such cases, and since it is derived from the central limit theorem it requires large samples in order to provide reliable approximate confidence intervals. This paper considers an alternative method of constructing approximate confidence intervals for the parameters of a simple linear regression model with Cauchy errors which is based the normal approximation to the Cauchy likelihood. The normal approximation to the Cauchy likelihood is obtained by a Tailor series expansion of the Cauchy log-likelihood function about the maximum likelihood estimate of the parameters and ignoring terms of order greater two. The maximized relative log-likelihood function for each parameter is then derived from the normal Cauchy relative log-likelihood function. The approximate confidence intervals for the parameters are constructed from their respective maximized relative log-likelihood functions. These confidence intervals have closed form confidence limits, are short and have coverage probabilities close to the nominal value 0.95.
Abstract: The estimators of the slope and the intercept of simple linear regression model with normal errors are normally distributed and their exact confidence intervals are constructed using the t-distribution. However, when the normality assumption is not fulfilled, it is not possible to obtain exact confidence intervals. The Wald method of interval estim...
Show More