Hello sir,
Plzzz tell me how can i get real values from my MLP. As MLP distinguish between only classes. can you please tell me how can i use MLP house price prediction challenge and get the real value in the output.
Neural Network Challenge-: House price prediction
hey @sambher_shubham ,
you just need to change activation function in final output layer to linear or tanh based on your predictions.
and along with it change your model loss to MSE.
It will work like regression task.
I hope this helps you .
Thank You. 
.
But bhaiya the range of tanh is from -1 to 1. So how can i use tanh? It is somewhat similar to sigmoid
You can use tanh , there is no restriction.
But i have experimented it many times it had always given poor results.
and can you just imagine house price to negative ,no naa…
so how can we use this in output layer.
So linear i think so is the best one?
And also one doubt how can i convert categorical data to numerical one?
Just encode them with classes as 0,1,2,3,4 ,etc.
You can use LabelEncoder for this purpose , it will do this task easily.
Its giving me error bad input shape as the shape is (1100,81)
can you please provide link to your code.
It will be easier for me to understand
i am giving you the link wait bhaiya
yeah sure. No problem.
I am here.
can i give you my python file?
yeah no problem. Just provide with your full code.
I have send the SS as i am using the discussion form for the first time so i am facing difficulty to send the file
LabelEncoder Just takes input as 1 column values , not the whole dataframe .
You need to use it only on those columns where you need it.
That i know but there are so many columns of categorical data
Applying one by one to every column will take time.
Try this :
df.select_dtypes("object").columns
it will return all those columns which have string values or categorical data which requires conversion.