Abstract With the popularity of the Internet and crowdsourcing, it becomes easier to obtain labeled data for specific problems. Therefore, learning from data labeled by multiple annotators has become a common scenario these days. Since annotators have different expertise, labels acquired from them might not be perfectly accurate. This paper derives an optimization framework to
Continue Reading “A Classification model for Diverse and Noisy Labelers”
We take one of the machine learning task from Kaggle, house price prediction, as the topic of this final project. One of the main reasons is that the provided dataset is relatively easy to parse and manipulate. It turns out that we have much time to do a thorough feature analysis for both numerical and
In this homework, I am working on the relationships between primal problems and dual problems, which is derived by solving Lagrange Multiplier. Among textbook problems, there are two coding problems that need to be solved with cvx package in MATLAB. Experiment results and codes are attached in the report.
The first part of this homework is problems extracted from textbook. They are mainly about log-concavity of Gaussian cumulative function, solving optimization problems with constraints, simplification of general optimization problems to specific quadratic problems, and generalized posynomials and geometric programming. Later on, for the second of the homework, I need to solve a predefined maximum
For the first part of this homework, I am asked some questions about Kullback-Leibler divergence and the differentiation between convex, concave, quasiconvex, and quasi concave functions. For the second part, I need to derive the update for log-sum-exp formula and implement a simple function to simulate Least Action principle in Physics with cvx package in
In the very first homework of this class, I first work on textbook problems, such as Voronoi sets and polyhedral decomposition, positive semidefinite cones, and properties of dual cones. Later, I am asked to find out the minimum polyhedron that can cover a sphere that is centered at (r, r, r) with radius r.
In this quarter project, we are asked to do a detailed analysis for an operating system in different aspects. For CPU and scheduling, we need to estimate the overheads for procedure calls and system calls, task creation time, and context switch time. For memory, we have to calculate Random Access Memory (RAM) access time, RAM
For the last homework, I am asked about not only the motivation of using scheduler activations, but also some details about Log-Structured File System (LFS), including what the assumptions LFS have and how data reading, writing, and segment cleaning are done in LFS.
In the first part of the homework, I elaborate the implementation details of Global Memory System (GMS). Later on, I am asked to compare the tradeoff between transparency and optimization across a few operating systems, including Sprite, Xen, and Grapevine. At the end, I compare the lazy evaluation policy between Mach, Exokernel, Xen, and VAX/VMS.
There are three parts in this homework. In the first part, I need to compare the details between different systems, including Hydra, Multics, Unix, and Pilot. Last, I am asked to provide an operating system example that takes advantage of specialized hardware. Lastly, the question is about the reason for existence of multi-user timesharing system,
Recent Comments