Logistic Regression Model. Week 10

(ordinary least squares) model is used when there is a binary dependent variable, the following problems might arise:
The error terms are heteroskedastic.
e is not normally distributed because Y takes on only two values;
The predicted probabilities can be greater than 1 or less than 0.

(aka “ln”)

estimating the coefficients of a logistic model.
Likelihood function:
The estimates b0  and b1  are chosen to maximize  this likelihood function.
Learn more about MLE:
https://medium.com/codex/logistic-regression-and-maximum-likelihood-estimation-function-5d8d998245f9

= titanic, family = "binomial")
summary(logit.m)

on how adults save, borrow, make payments, and manage risk. The data are collected in partnership with Gallup, Inc., through nationally representative surveys of more than 150,000 adults in over 140 economies. ”findex_uzb.csv” provides some of the variables from Findex dataset of Uzbekistan for 2017. Responses are recorded as 1 if “yes”, 0 if “no”.
There are total of 1000 observations in the dataset.
Fit a simple logit model to predict whether the person saved (1) or not saved (0) based on age.
Comment on the significance of Age.
b) Predict the probability of saving at a financial institution for a person aged 45 years old.
c) Fit a multiple logit model to predict whether the person saved (1) or not saved (0) based on age, gender, education level, and employment status.
d) Compute the odds ratio and probability for a female, 30 years old, with secondary education, and employed.

in R).
ISBN: 978-1-4614-7137-0. p.p. 156 – 160.
2. https://www.datacamp.com/community/tutorials/logistic-regression-R
3. https://machinelearningmastery.com/logistic-regression-for-machine-learning/