(Closed)SSL&SS Class Information(2024)

SSL: Class Schedule (課程為實作取向, 確認大家都會具備相關實作的能力)

6/28(五) – Week 1
10:00-10:55 – CI/CD (FH)
14:00-15:25 – NN Fundamentals (FH)
15:30-17:00 – Back Propagation (JJ/JW)

7/5(五) – Week 2
10:00-12:00 – Overfitting and Underfitting (JJ)
14:00-15:25 – Skills for Comating Overfitting (JJ)
15:30-16:25 – CNN Programming (JW)
16:30-17:30 – API Development (Wayne)

7/12(五) – Week 3
10:00-12:00 – YOLOv8 and Later Versions (Wayne)
14:00-15:00 – FH-vGPU (FH)

7/19(五) – Week 4
10:00-12:00 – Telegrambot Programming (Wayne/YH)
14:00-15:30 – RL Programming (JW)

SS: Class Schedule

7/22(一)
10:00-12:00 – Concepts of Deep Learning(1/2). (FH)
14:00-16:00 – Concepts of Deep Learning(2/2). (FH)

7/23(二)
10:00-12:00 – Practical Issue of Model Training – Overfitting and Underfitting. (FH)
14:00-17:00 – CNN Structure and Its Technical Operations. (FH)

7/29(一) 7/24(三) (因颱風延期)
10:00-12:30 – The Variants of CNN Models. (FH)
14:30-17:00 – The Introduction of CNN Object Detection Models. (FH)

7/30(二) 7/25(四)
10:00-12:00 – Lightweighting Methods of CNN. (FH)
14:00-17:00 – VAE: Crafting Realities with Probabilistic Distribution. (PW)

7/31(三) 7/26(五)
10:00-12:30 – GAN: Image Generation Through Competition (PW)
14:30-17:00 – Diffusion Models: Sculpting Order from Chaos. (PW)

8/1(四) 7/29(一)
10:00-12:30 – Introduction of Transformer and ViT (PW)
14:30-17:00 – The Cutting Edge Sequential Model – Mamba. (PW)

8/2(五) 7/30(二)
10:00-12:30 – A Survey of Prompt Engineering. (PW)
14:30-17:00 – Introduction of Federated Learning. (PW)

8/5(一) 7/31(三)
10:00-12:30 – Introduction of Reinforcement Learning. (MH)
14:30-17:00 – Variants of Reinforcement Learning Algorithms. (MH)

8/6(二) 8/1(四)
10:00-12:30 – A Survey of SOTA Reinforcement Learning Models (MH)

HWs: (Turn in your reports in LeTeX.)

1. (for all SSL members – {JW}) Given compperf-db-backup (which works in MySQL DB) in LX4-laptop. Currently, it is found that certain slow queries are occurred in the computer-performance-monitoring website. Identify the slow queries and solve them. (Due: 7/30) (Working platform: FH (your PC), MH (250), JJ (JJ’s RPi), Wayne (your PC), TT(LX4))
   (1) List the queries, together with their response time, used in the website.
   (2) Analyze the skills in shortening response time for the slow queries: (i) partitioning tables, (ii) creating index, (iii) both of them. And provide experimental comparisons.
   (3) Give your suggestions for solving the performance bottlenecks to the website.
   Ref. for HW1:
   https://github.com/major/MySQLTuner-perl
   Google: mysql tuning, script, 2024
2. (for JJ,Wayne,TT) We have an utp-ftpserver and a utp-ftpclient which are developed by FH almost two years ago for collecting images between PTH-RPi and api-server under very low-bandwidth network. (Due: 8/31)
   (1) Analyze how this program work. Also, give three drawbacks and your improvements.
   (2) In case we have an extra HDD, how do you move the pig data directory to the new HDD? Explain your idea and give the detailed steps. (Wayne and TT will move the directory in Sep.)

Member	HW1(7/30)	HW2(8/31)
FH	x	x
JW	x	x
MH	x	x
JJ	missing assignment	Failed
Wayne	v	Failed
TT	v	Failed

設備分配:

JJ: PC, RPi

MH: PC

JW: PC

YH/Wayne Team: PC+RPi+laptops