ME 697Y Assignment Problem Sets (Spring 2024)

Note: All files are in Word or pdf format.

Syllabus:

Project 1: (in-class presentation Feb. 13 & 15): Choose the one that has not been selected after an announcement is made in class.

You need to submit your short proposal by Feb. 6 (the earlier, the better) and upload the final presentation slides (in power point) and programs in one zipped file by 2:00pm on Feb. 12 into the Brightspace Assignment Project 1.

1. ANFIS: Elias Pergantis, Andrew Manion
2. FFNN using BP training: Hanzhi Yang
3. FFNN for a dynamic system using BP training: Andrew Manion
4. FFNN using GA:  Juhyung Kim
5. RBFN using LS: Elias Pergantis
6. RBFN using OLS: Oscar Yu
7. RBFN using OLSGA:
8. FBFN using ALS:  Jackson Cruise
9. FBFN using OLS: Alejandro Coss
10. Recurrent neural networks or back propagation in time: Will Farlessyost
11. CNN: Michael Goldberg
12. Any other NN paradigm:

Topic 1 and 5 must be done together.

Project 2: (in-class presentation on March 5 & 7): 

You need to submit your short proposal by Feb. 27 (the earlier, the better) and upload the final presentation slides (in power point) and programs in one zipped file by 2:00pm on March 4 into the Brightspace Assignment Project 2.

• Genetic algorithm: Elias Pergantis
• Genetic algorithm for mixed integer problem:  Alejandro Coss
• Evolutionary Strategies (mu, lambda):   Oscar Yu
• Evolutionary Strategies (mu + lambda):  Will Farlessyost,  Jackson Cruise
• Extended ES:
• PSO: Hanzhi Yang
• Support Vector Machines: Michael Goldberg
• Comparison between two optimization methods: Andrew Manion, Juhyung Kim

Project 3: (in-class presentation on April 2&4): You will need to submit your proposal by March 26 and upload the final presentation slides (in power point) and programs in one zipped file by 2:00pm on April 1 into the Brightspace Assignment Project 3.

• Model reference control using neural networks: Oscar Yu
• Inverse control using neural networks and fuzzy logic:
• Model predictive control using neural (or fuzzy) networks: Elias Pergantis
• Self tuning adaptive control using neural networks: Juhyung Kim
• Fuzzy PID supervisory control of robot arms: Michael Goldberg
• Fuzzy PID supervisory control: Will Farlessyost
• Fuzzy PID control: Andrew Manion
• Hybrid fuzzy control: 
• Fuzzy model reference learning control: 
• Fuzzy internal model control:  Jackson Cruise
• Fuzzy model optimal control: Hanzhi Yang
• Self organizing fuzzy PID control:  Alejandro Coss

Presentation Outline

1. Problem definition (with mathematical function or real data)

2. Training results

3. Test results

4. Graphical approximation results (compared with the original function) if possible

5. Conclusions

Items to be submitted

1. Presentation slides (power point slides)

2. Programs (in zipped format) with the instruction on how to run the programs

Final Term Project

1. Proposal due date:  March 28, 2024 (download the guideline).  The latest deadline is April 4.
2. Topic: your choice (can be an extension of earlier work, combination of multiple, a new one)
3. In-class presentation:  April 23 and 25 (15 min per person)
4. Upload the final presentation slides (in power point) by 2:00pm on April 22 into the Brightspace Assignment Final Project.
5. Due date (program and final report):  upload the final report (in word and pdf) and programs in one zipped file by 5:00 pm on April 26.

See the guideline of the final report.