New to Edge Impulse: Need Guidance for Dataset Impulse Creation

Help
#1

Hello,
I have a dataset with 5 columns - timestamp, temperature, humidity, light intensity, and temperature in shadow. I aim to predict the temperature in shadow by training a machine learning model. Can you please advise me on the best model that suits my dataset? Also, I am new to using Edge Impulse and need help creating an impulse from this data. Can you kindly provide detailed steps for me to follow?

Your assistance is much appreciated.

#2

Hi @AliSalman

Welcome to Edge Impulse, we have a large set of guides and blogs for this. Let me get a good starting point for you here:

Once you get comfortable with data acquisition you can move on to more involved methods:

Best

Eoin

#3

Hello @AliSalman,

So in your case, the temperature in shadow should be the label.
What you can do if to name your csv such as:

20.your-file-name.csv when your temperature in shadow is 20 and remove that column from this CSV file.
Also note that your timestamp should have a constant increase.
e.g. with 200ms:

timestamp, column1, column2, column3,
  0, x, y, z,
200, x, y, z,
400, x, y, z,
...

Best,

Louis