How to avoid overfitting with linear regression ? Give at least two solutions and explain your rationale
How to avoid overfitting with linear regression ? Give at least two solutions and explain your...
a) What is overfitting problem? How does regularization solve the overfitting problem? Explain Explain with example. [Hint: Ridge regression] (3 + 5 marks) b) What is logistic function? Why do you need to use logistic function in linear regression? Explain with example. (2 + 5 marks) c) Explain the concept of bias-variance trade-off. What will be the effect on bias if we regularize the weights in linear/logistic regression model? Explain in brief. (5 + 5 marks)
5. In your own words, explain why overfitting and underfitting are not desirable? How would you confirm that your model is overfitting? State two methods to combat model overfitting. (15)
5. In your own words, explain why overfitting and underfitting are not desirable? How would you confirm that your model is overfitting? State two methods to combat model overfitting. (15)
Q1 a) Explain what it means that the ordinary least squares regression estimator is a linear estimator, paying specific attention to how it implies independent variables interact with each other. b) Give two examples of models where the parameters of interest cannot be directly estimated using OLS regression because of nonlinear relationships between them. c) What is the minimum set of conditions necessary for the OLS estimator to be the most efficient unbiased estimator (BLUE) of a parameter? List each...
Give an example of an Integer Linear program which has no feasible integer solutions, but its LP relaxation has a feasible set in R2 of area at least 10
Give an example of an Integer Linear program which has no feasible integer solutions, but its LP relaxation has a feasible set in R2 of area at least 10
Briefly explain two ways to limit overfitting in constructing a decision tree. Briefly explain the advantages and the weaknesses of decision trees.
Explain the elements of a regression equation for a simple linear regression: Y=b+mx. Why are regression analysis useful? Give an example.
What are the pitfalls of simple linear regression? True or False for each Lacking an awareness of the assumptions of least squares regression. Not knowing how to evaluate the assumptions of least squares regressions. Not knowing the alternatives to least squares regression if a particular assumption is violated. Using a regression model without knowledge of the subject matter. Extrapolating outside the relevant range of the X and Y variables. Concluding that a significant relationship identified always reflects a cause-and-effect relationship.
Give two reasons that may account for a low R squared value of a linear regression model
Explain how a business could use a scatter plot and linear regression to develop a model for the business and what the rate of change would mean.
Explain how a business could use a scatter plot and linear regression to develop a model for the business and what the rate of change would mean.
Give examples of at least two market failures and explain how they represent challenges to the free market understanding of business's environmental responsibilities.