Scribd
Upload
3 views2 pages

Predictive Modeling Assignment

The document compares linear and logistic regression, highlighting their different purposes and output types. It discusses the advantages of decision trees for complex relationships and the benefits of random forests over single decision trees. Additionally, it explains the differences between R² and accuracy, and outlines preprocessing steps necessary before building predictive models in Scikit-Learn.

Uploaded by

dkamonde01
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views2 pages

Predictive Modeling Assignment

The document compares linear and logistic regression, highlighting their different purposes and output types. It discusses the advantages of decision trees for complex relationships and the benefits of random forests over single decision trees. Additionally, it explains the differences between R² and accuracy, and outlines preprocessing steps necessary before building predictive models in Scikit-Learn.

Uploaded by

dkamonde01
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd

1.

Comparison of Linear Regression and Logistic Regression


Linear regression and logistic regression are commonly used predictive models in data
analytics. However, they are used for different types of prediction problems depending on
the nature of the outcome variable.

Aspect Linear Regression Logistic Regression

Purpose Predict continuous Predict categorical


numerical values outcomes

Dependent Variable Continuous values such as Binary or categorical


revenue or sales outcomes

Output Numeric prediction Probability between 0 and


1

Model Function Linear equation Logistic (sigmoid) function

Typical Use Cases Sales forecasting, demand Customer churn prediction,


prediction fraud detection

2. Business Scenario Where a Decision Tree is Preferred


A decision tree is often preferred when the relationship between variables is complex or
non-linear and when the model must be easily interpretable.

For example, a bank deciding whether to approve or reject loan applications may use a
decision tree. The model can consider factors such as income, credit score, employment
status, and existing debts. The tree structure clearly shows the decision rules used to
approve or reject loans.

This transparency makes decision trees easier for managers and stakeholders to
understand compared to models like logistic regression.

3. Advantages of Random Forest Over a Single Decision Tree


• Higher prediction accuracy due to combining results from multiple trees.
• Reduced overfitting because predictions are averaged across many models.
• Better performance with large datasets and many variables.
• Ability to measure feature importance.
4. Difference Between R² and Accuracy
Metric R² (R-Squared) Accuracy

Used For Regression models Classification models

Meaning Explains how much Percentage of correct


variance is captured by the predictions
model

Value Range 0 to 1 (sometimes 0 to 1


negative)

Interpretation Higher value indicates Higher value indicates


better model fit better prediction
performance

5. Preprocessing Steps Before Building Predictive Models in Scikit-


Learn
• Handling missing data by removing rows or filling values using mean, median, or
mode.
• Encoding categorical variables using techniques such as label encoding or one-hot
encoding.
• Feature scaling using methods such as StandardScaler or MinMaxScaler.
• Splitting the dataset into training and testing sets.
• Detecting and removing outliers that may affect model performance.

You might also like