missing_treatments: Treats missing values in a given dataset using the best...

In akunuriYoshitha/MissingTreatment:

Description

missing_treatments initially eliminates all the columns having more than 70

• Zero Imputation

• Min Imputation

• Max Imputation

• Mean Imputation

• Median Imputation

• Mode Imputation

• Multiple Imputation - Random Forest

• Multiple Imputation - Predictive Mean Matching

Of all these techniques, Best technique is chosed based on the performance of different datasets on a linear/logistic model.

Usage

missing_treatments(data, dv)

Arguments

data	Any data frame that has to treated for missing data
dv	Dependent variable in the given dataset in order to eliminate that field while treating missing data

Details

Takes in a data frame and performs the best possible missing value treatment to each of the columns in the data frame

Value

Returns a list of 3 objects:

missing_treated

Missing value treated Dataset

fit_model

A list of model fit files for each of the techniques used

model_perf_metrics

Performance metrics for eac of the intermediate datasets created

akunuriYoshitha/MissingTreatment documentation built on Dec. 31, 2020, 7:40 p.m.