S = x1 = 1 10000 1 10 - - 1 10000 W1 w₁ W1 + 1000 W3 + 100 5 + 1/10 7 W5 - W2 = x2 = 100 W4 = 95 49 - X3 = W6 = 1 X4 = W8 = 1000 499 50 10000 4999 500 - - - 15 1000 wi W1 ພາ 10000 W1 1 500 W1 - - - - - - 100 w3 W3 W3 W3 1000 W3 50 W3 - - - - 100 15 W5 W5 W5 ພຣ - - W7 W7

We use machine learning to solve problems if they satisfy three conditions. What are these conditions?

[a] Prove that the "Bin Packing Problem" (BPP) is NP-Complete. Like the rest of the exam, you may use online resources but should cite your source and summarize the proof in your own words. [b] Write an approximation algorithm for the online version of BPP that randomly considers up to three partially filled bins, and if none have room, then create a new bin and use that. [c] What modification to your solution in [b] would you recommend to solve the offline version of BPP. [d] In light of your prior knowledge of and experience with approximation algorithms for BPP, provide a reasonable estimate of the error bound

Do we need to make iteration even if the team is one member