Skip to content

Latest commit

 

History

History
30 lines (18 loc) · 780 Bytes

assignment_q8_subjective_answers.md

File metadata and controls

30 lines (18 loc) · 780 Bytes

ES654-2020 Assignment 3

Your Name - Your Roll Number


Write the answers for the subjective questions here

Theoretical time complexity for normal method is O(P^2 * N) + O(D^3) Theoretical time complexity for gradient descent method is O((t+N) * D^2)

  • For normal equation, for a constant N, as P increases, there is huge increase in time computation as seen in heat map

  • The time complexity increases as a product of number of samples and number of dimensions while keeping number of iterations constant.

Heatmaps

For gradient descent

For Normal method

Using matplotlib

Varying N

Varying P