Compute posterior: P (C2 | x)

Exokernel 1. [4 points] A library OS “B”, on top of Exokernel experiences memory pressure and needs additional physical memory. Exokernel decides to revoke 20 physical page frames from library OS “A” to service this request. Describe how Exokernel can revoke the physical page frames from library OS “A” and how library OS “A” will respond.

Spin The context for this question is the same as the previous question. [6 points] Your friend — having not read a research paper in 3 years — is struggling to understand the specific benefits SPIN provided in the “Extensibility, Safety and Performance in the SPIN Operating System” paper that was assigned. He was hoping you could briefly describe (in a sentence or two) each attribute listed in title of the paper, since the paper’s title was the only part that really stuck with him. b) [2 points] How is SPIN safe? 

Paravirtualization [5 points]  You are launching an application inside a para virtualized VM. Your expectation is that your application will have memory isolation and protection from other applications on the same VM.  With succinct bullets, explain how this expectation is fulfilled. 

Paravirtualization Above picture shows the I/O ring data structure used in Xen to facilitate communication between the guest OS and Xen.  Guest-OS places a request in the I/O ring using the “Request Producer” pointer. Xen places a response in the I/O ring using the “Response Producer” pointer.  [1 point] A guest OS is polling the ring. What exact condition tells it that a response is available to consume, and what simple step should it take before reading the descriptor?

You want to check if your convolutional network can be affected by adversarial examples – list two methods you could perform to check. Answer using up to 3 sentences max.

Section 4 – Bayesian Learning

Write pseudocode to implement the ROI pooling forward pass logic for a single 2D data sample (no batches). Clearly mark your inputs, your outputs, and your parameters. No need to handle special edge cases. Using numpy calls/syntax is allowed. Points may be deducted for lengthy solutions. roi_pooling_forward() { return }

Section 1 – Randomized Optimization

A PTA measures a patient’s knee flexion before and after a 4-week exercise program. This measurement is MOST likely obtained using which tool and for what purpose?