There are 3 repositories under ntu topic.
台大 計算機安全 - Pwn 簡報、影片、作業題目與解法 - Computer Security Fall 2019 @ CSIE NTU Taiwan
Machine Learning and having it Deep and Structured (MLDS) in 2018 spring
This is a course-downloader to help NTU students download courses data from NTU Ceiba.
Thesis Latex Template for Nanyang Technological University (NTU)
National Taiwan University Master's Thesis Template (Latex + Word)
🌆 🏙 🌃 Viz.js Graphviz - An Elegant Visualizer for And-Inverter Graph
NTU Computer Science Tutorials, Labs and Assignments
🎓 Unofficial LaTeX templates for your graduate thesis (both master's theses and doctoral dissertations) at National Taiwan University. 國立臺灣大學碩博士學位論文 LaTeX 模板
Incubated Judge-Girl's (A Cloud-Native Online Judge System) root project. With powerful CMS features (e.g., dry-run / compilation-script generation / Continuous Problem Integration & Delivery) and educational features like exam and assignments. Judge Girl is carefully designed of its plugin-like structure easily supporting various judge-policies and code-quality inspectors in the future.
Hung-Yi Lee ML 2017 Spring Homework
🎒 Materials (slides, handouts, homework, cheat sheets...) of the courses I took or audited at National Taiwan University.
The frontend of NTUCourse Neo.
EEE Latex Dissertation Template, Nanyang Technological University (NTU), re-defined in 2021
Discrete Hidden Markov Model (HMM) Implementation in C++
Applied Deep Learning (2019 Spring) @ NTU
📆 Powerful and elegant university timetable builder (NUS and NTU)
Pytorch implementation of CS-Tacotron, a code-switching speech synthesis end-to-end generative TTS model.
Backend API service for NTUCourse-Neo.
台大碩博士論文模板 (R Package)
Youtube李宏毅教授(Hung-yi Lee, NTU)讲解的《Machine Learning》 课程笔记与notebook相关实现。
A Python project focused on implementing and comparing various algorithms for customer service staff scheduling automation achieved through operations research, with a particular emphasis on a new algorithm proposed by the authors.
🤖 An automated NTU Thesis LaTeX continuous integration and continuous deploying service built up with GitHub Actions.
Swarming behaviour is based on aggregation of simple drones exhibiting basic instinctive reactions to stimuli. However, to achieve overall balanced/interesting behaviour the relative importance of these instincts, as well their internal parameters, must be tuned. In this project, you will learn how to apply Genetic Programming as means of such tuning, and attempt to achieve a series of non-trivial swarm-level behaviours.
Hung-Yi Lee Linear Algebra 2018 Fall Homework
In this lab, you are required to complete a virtual memory implementation, including how to get a physical frame for a virtual page from the IPT if it exists there, how to put a physical frame/virtual page entry into TLB, and how to implement a least recently used page replacement algorithm. A software-managed TLB is implemented in Nachos. There is one TLB per machine. There is also an IPT which maps physical frames to virtual pages. Basically, the translation process first examines the TLB to see if there is a match. If so, the matching entry in the TLB will be used for address translation. If there is a miss, the IPT will be looked up. If a matching entry is found in the IPT, the entry will be used to update the TLB. A miss in the IPT means that the page will have to be loaded from disk, and a page in and page out will be performed. To decide which page to page out, a page replacement policy is used, for example, a least recently used algorithm which will be explained late on. During each lookup process, you need to perform some checking in order to make sure that you are looking up the correct entry and that the entry is valid. In order to check whether you are referencing the correct entries from the TLB, you have to check the valid bit. The TLB will get updated when an exception is raised and the required page entry isn't in it. In this case, a new entry needs to be inserted into the TLB. The new entry will be inserted into an invalid entry in the TLB or replace an existing entry if it is full. Since the TLB is small, the replacement policy for the TLB is simply FIFO. When there is a context switch between processes, e.g. the main process executing a child process, the entries in the TLB will be cleared by setting all entries to invalid. The IPT is simply implemented using an array, represented by the memoryTable (a mapping of what pages are in memory and their properties). There is one entry for each of the physical frame, and each entry contains the corresponding process id, virtual page number, and the last used field that records the tick value when the page was last accessed. The least recently used algorithms works by iterating through the memoryTable, from the beginning, to look for the entry that has been least recently used. If there is an entry that is not valid (i.e., its process is dead), the algorithm will return the index of this invalid entry. Otherwise, the algorithm will return the index of the least recently used entry (that is, the entry with the smallest last used field).
The complete course reviews for SCSE Computer Engineering students and student-to-bes