Steps In Regression Analysis

In regression analysis, model building is a process of developing probabilistic model that describes the relationship between dependent and independent variables.
In general, model building process includes different steps such as collecting of data, understanding and paying attention to what extent the data important, finding the statistical, mathematical or simulation model to make predictions. Model building is a crucial skill to acquire in every filed of science.
Steps involved in Model Building
The general steps involved in Model building are as follows Figure 2.2.1: Steps for Model Building

Step 1: Prepare the data the marketing manager in an organisation solves the problems confronted in the field of marketing.

Step 2: Optionally

It uses historical data to predict a categorical target. This function will be even used for anomaly detection. In this case, the SVM algorithm with a null target is used (One-Class SVM).
Clustering Is a descriptive mining function. It identifies natural groupings within a data set.
Clustering models can use: k-Means or O-Cluster algorithms. The default is k-Means.
Feature_Extraction Is a descriptive mining function. It creates an optimized data set on which to base a model.
Feature extraction models use the Non-Negative Matrix Factorization algorithm.
Regression Is a predictive mining function. It uses historical data to predict a numerical target.
Regression models can use Support Vector Machine or Linear Regression. The default is Support Vector Machine.

Step 4: Model Details
This step explains in detail about attributes of the model and rules of the model. To retrieve model details you can invoke GET_MODEL_DETAILS and a separate GET_MODEL_DETAILS function is existed for each algorithm.
Model details will reverse the transformations applied by the attributes. Thus the information is easily understood by the user. By invoking the GET_MODEL_TRANSFORMATIONS functions, you can obtain the transformations embedded in the

Classification Model: A classification model can be tested by applying it to test data with known values, even compares the predicted values with known values.
Regression Model: Even the regression model can be tested by applying to the test data with known target values, compares the predicted values with known values.
For both classification and regression the test data must be compatible with the data, which is used to build a model. The build data and the test data will come from the same historical data set. A percentage of records is used to build a model, the remaining records are used to test the model.
These test metrics are used to assess how accurately these models predicts the known values. If these models performs well and meets all the business requirements, then it can be applied to the new data to predict the future.
Step 6: Evaluate the model
Evaluation of the model can be done with some