How to Predict metal temperature by using the Machine learning ?

I want to predict the metal temeperature by using some other 3-4 independant parameters. So respone variable is Metak temperature. Its should consider dynamics and thermal inertial. Which ML algorithum or method should I use ?

Answer 1

[Mahesh Taparia]

Hi

This is a regression problem. There are many regression algorithms like linear regression, SVR, Neural networks, etc which can be used.

You can refer to this documentation for details of those algorithms in MATLAB. Moreover, there is an app called Regression Learner, you can also use that. For more details, you can refer to this documentation of that app.

For application of neural network, you can refer to this documentation of train function. Hope it will help!

[Koustubh Shirke]

Hi Mahesh,

Thanks for the answer. As per my understanding, all these algorithm does the statis modelling and predictions. I want to have dynamic model/predictions where current state will be dependant on the previous one step. FOr example, while predicting the metl temperatures, it should cosider the thermal lag which has been coming from prevuios stapes.

Regards,

Koustubh

[Mahesh Taparia]

Hi

So, the output is time dependent. You can try with LSTM (long short term memory) based approach. For this you can refer to this documentation. Moreover, you can use neural network time series modeling. For this, refer to this documentation.