The logit and fitting analysis of logistic regression on jakarta composite index in covid-19 pandemic

mutijah, mutijah (2020) The logit and fitting analysis of logistic regression on jakarta composite index in covid-19 pandemic. Faculty of Mathematics and Natural Science International Seminar on Innovation in Mathematics and Mathematics Education (ISIMMED). pp. 1-6.

Preview

Text Full Paper-Mutijah.pdf Download (238kB) | Preview

Abstract

Many statistic study about forecasting, predictive, estimation, or relationship have been developed globally either before or during the covid-19, including in Indonesia. This study purposes to analyze the odds and fitting of the relationship between the increasing and decreasing daily of covid-19 cases with the up and down of Jakarta Composite Index. Logistic regression method was used to analyze them. Logistic regression, also called logit regression or logic model was one of the methods to analyze in determining of the relationship between two or more variables which the dependent variable is category and the independent variables can be categorical, continuous, or a combination. The relationship can be in variables between economy and physics, economy and agriculture or biology and medical,etc. The odds played an important role in logistic regression. In this study, the odds was a good indicator that showed the opportunity of up against down under the data of Jakarta Composite Index. The analysis were done on variable of the increasing and decreasing daily of covid-19 cases in continuous and categorical. This analysis were based on Jakarta Composite Index historical data for 120 days on March 17,2020 until September 16, 2020. The analysis results were obtained that the odds established an up was 1,267 times as likely as a down in continuous independent variable and 1,264 times in categorical. While the probability established the logit model was not fit with the values of goodness-of-fit very low. This matter did not mean that the logic model was made to produce a false regression. The value of goodness-of-fit on variable of the increasing and decreasing daily of covid-19 cases in continuous is greater than in categorical

Actions (login required)

View Item