DAT-430: Module 5 Assignment
Southern New Hampshire University
When completing data analysis, there are many options for what type of modeling
technique you wish to use to predict outcomes or visualize relationships between variables
, within a dataset. We will be looking at seven different models, briefly discussing their strengths
and identifying use cases for each one.
The first data model we will be discussing is the ordinary least squares (OLS). This technique is
used to estimate the coefficients of linear regression equations (XLSTAT, 2017) – these
equations describe the relationship between and independent variable and one or more
dependent variables. This technique is especially useful when trying to predict the outcome of
something based on interactions with other factors – such as in meteorology when predicting
the chance of rain based other environmental factors like temperature and humidity.
Generalized linear models (GLM) are a type of regression model that, similar
to linear regression models, depicts a relationship between a response
variable and predictor variables. The GLM model, however, does not assume
a linear relationship between the variables, which allows for more flexible
modeling of complicated relationships. The flexibility of a GLM allows for
modeling linear relationships, Poisson Distributions and Binomial
Distributions.
Logistic Regression is a data modeling technique that identifies the
probability of a specific outcome when given an input variable. This type
of
11:20:07 -06:00
Southern New Hampshire University
When completing data analysis, there are many options for what type of modeling
technique you wish to use to predict outcomes or visualize relationships between variables
, within a dataset. We will be looking at seven different models, briefly discussing their strengths
and identifying use cases for each one.
The first data model we will be discussing is the ordinary least squares (OLS). This technique is
used to estimate the coefficients of linear regression equations (XLSTAT, 2017) – these
equations describe the relationship between and independent variable and one or more
dependent variables. This technique is especially useful when trying to predict the outcome of
something based on interactions with other factors – such as in meteorology when predicting
the chance of rain based other environmental factors like temperature and humidity.
Generalized linear models (GLM) are a type of regression model that, similar
to linear regression models, depicts a relationship between a response
variable and predictor variables. The GLM model, however, does not assume
a linear relationship between the variables, which allows for more flexible
modeling of complicated relationships. The flexibility of a GLM allows for
modeling linear relationships, Poisson Distributions and Binomial
Distributions.
Logistic Regression is a data modeling technique that identifies the
probability of a specific outcome when given an input variable. This type
of
11:20:07 -06:00
