Description
you may select your own database to use just be sure to add the file part of the submission folder. The code show be displayed in pdf format once completed showing both the code and the codes various outputs.
Unformatted Attachment Preview
Homework Assignment 3 – CAP 4773 – Intro to Data Science
Due 11-07-23
Submission Guidelines:
Submit your report on Canvas in PDF format.
Code: Ensure all your code is well-commented. Each section should be clearly demarcated with
comments.
Visualization: All plots should have relevant titles, x and y-axis labels, and legends where necessary.
Analysis: Your comments on visual and numerical results should be based on the evidence from your
data.
Use the references from the class Jupiter notebooks and class examples to work and compile your
report.
Use a dataset different from those used in class examples or exercises.
Sections
Apply the following concepts to your dataset. Where possible, display statistics and graphs to elucidate
your findings. Keep your explanations simple and straightforward.
Section 1 – LDA (20 points)
Use Linear Discriminant Analysis (LDA).
Create classes based on existing features, if necessary.
Section 2 – K-fold Cross-validation (15 points)
Implement K-fold cross-validation to assess the model’s performance.
Section 3 – Linear Model Selection Statistics (20 points)
Evaluate the model using the following statistics:
Cp
AIC
Adjusted R^2
BIC
Section 4 – Lasso and Ridge Regularization (15 points)
Apply Lasso and Ridge regularization techniques and compare their effects on the model.
Section 5 – Simple Polynomial Regression (15 points)
Implement a simple polynomial regression on your dataset.
Section 6 – Spline Polynomial Regression (15 points)
Apply spline polynomial regression and observe the differences compared to the simple polynomial
regression.
Section 7 – Bonus (20 points)
Based on your findings from Sections 1 through 6, write a concise paragraph detailing conclusions about
your dataset. Share your insights with the class in the provided discussion forum.
Purchase answer to see full
attachment
