This means that you should come up with an equation for your model. The equation should be written using the z-domain syntax and can be simple polynomials or rational functions (depending upon the technique being used). The output is one of the 11 classes (words) that your model βsomehowβ detects.
Methodology
The data is saved in 11 folders (according to the words themselves) here (Data). You should tentatively follow the methodology given below:
1. From each folder, pick-up 65% data (waveforms) randomly. {MATLABβs randi function can help}
2. Use any technique(s) to build your model using this 60% of this data subset while using the remaining 5% for validation of your model as the model building process takes place.
4. Once the model is finalized record the structure and coefficients and perform some statistical analysis. { https://www.youtube.com/watch?v=4ipdsefA5ik }
6. Present your results graphically as much as possible.
9. Report template will be uploaded separately.
10. Tentative Grading Rubric for Project work:
a. Projecthon 50%
b. Discussion 40%
c. Report 10%
Important Note
DO NOT USE CHAT-GPT!!!
All coding MUST be done in MATLAB (QU-cloud version ONLY) for consistency.
