ISSN: 1813-0410

Author : عباس عمران, نور


Use a bootstrap method in parametric regression models (Logistic Model)

journal of kerbala university, 2017, Volume 13, Issue 4, Pages 90-104

Health is a blessing from God many we have , which enables man to live a normal life and being able to enjoy his life , so since time immemorial has repeatedly rights to develop and search for healing treatment , and among the things that are working on all governments and all workers to health and development, and the international community to improve The development is the health care and patient care.
So according to this research study of the relationship between the care provided to the patient during the period of lying in the hospital and between smoking, sex, age and weight by simple random sample size pull (182) infected with stroke people from the al-Husseini hospital in the holy city of Karbala during the period ( 2014- 2015 ) using a specimen traditional logistic regression and bootstrap , identifying any of those factors ( smoking , sex , age , weight) the most impact on the patient's stay in hospital , and clarify the impact of the use of style Bootstrap on the form Teachers logistical measures and then compared with real data results with non-real ( generated ) in order to determine which one best use real data or non-real (generated).
And research findings that the use of style bootstrap gives more accurate results , and that the specimen logistic regression Bootstrap real data is the best specimen in the representation of the data.