Intelligent Computing Methods

Course Project

In the course work you will carry out a small project, using the techniques that you have

learnt, to tackle a practical technological problem. You will be asked to investigate the

learning algorithms such as BP$,$ RBF$,$ SOM, FIS, GA$,$ PSO and SVM$,$ etc. You will be

expected to develop your own algorithm to solve a real problem, using the packages

supplied in Matlab, Python or other languages, and to write a project report. Function

approximation: design a neural network to fit the following MIMO functions:

$y_1=2+x_1^{1\mathrm{.}5}-1\mathrm{.}5sin(3x_2)$;$y_2=x_{2}sin(x_1)+x_{1}cos(x_2)$;$1$

Generate $200$ data, $100$ points for training and the other $100$ for model validation

and testing.

You are encouraged $to$ use real engineering data $to$ solve real problems.

$In$ the work, you should take the following points into account:

Reproducibility

Don't claim things without references $or$ other evidence

Analysis $of$ results

More important $to$ clearly present your ideas $on$ how things should $be$ solved,

than $to$ actually have succeeded $(had$ time $to)$ solving them!

A failed experiment $is$ not a failed project! Failed experiments can $be$ analyzed

and learnt from $(and$ written about$)$

Suggest future work $\frac{?}{?}$ extensions

Submission: please submit your source code $(.$mat $or.py$ files; including user manual,

environment, etc.$)$

Note:

Source code $is$ needed for results reproduction.

The report can $be$ written $in$ English $or$ Chinese.

Please strictly follow the format $of$ IEEE Transactions.