Implement multi-task neural network from scratch in python
Budget: ₹600 – ₹1,500 INR
Implement a multi-task neural network from scratch (incl.
back-propagation) having the following attributes:
(i) 2 inputs (both continuous)
(ii) 2 hidden layers (one has sigmoid and another has tanh as the
activation function)
(iii) 2 outputs (one for binary classification and another for regression)
Please find the relevant dataset attached. Use stochastic gradient descent
approach and have 80:20 split for training and validation.
NOTE: For such multi-task learning problems, you can do a weighted addition
of the losses involved.
back-propagation) having the following attributes:
(i) 2 inputs (both continuous)
(ii) 2 hidden layers (one has sigmoid and another has tanh as the
activation function)
(iii) 2 outputs (one for binary classification and another for regression)
Please find the relevant dataset attached. Use stochastic gradient descent
approach and have 80:20 split for training and validation.
NOTE: For such multi-task learning problems, you can do a weighted addition
of the losses involved.