Review Questions

Class 1 Review

1. What happens when alpha, the learning rate is too large? Too small?

2. How can we combat overfitting our data?

3. True or False?: You must adjust the learning rate alpha throughout the course of your program.

4. How do you calculate the MSE (Mean Squared Error), or the cost function?

Class 2 Review

1. Qualitatively, how do we add two vectors? What about matrix addition?

2. What is the definition of a vector geometrically?

3. If the dimensions of two matrices are "m"x"n" and "p"x"q", then what are the dimensions of the product of these two matrices? What must be true in order for matrix multiplication to be defined?

Class 3 Review

1. Qualitatively, what is the transpose of a matrix?

2. Up to how many elements can we have in a vector?

3. What kind of Python variable names will not work? (Hint: There are 3 types.)

4. How do we take the elements at index 2,3,4, and 5 from a list using list slicing? (Hint: Remember, in list slicing, the first number determines the index of the starting element and the last number determines the index of the last element called. Note that the element at the index indicated by the last number is not called.)

5. How do we define a NumPy array?

6. How do we multiply the corresponding elements of two arrays? How do we multiply the two arrays through matrix multiplication? How are these two different?