Objective

• How to make a prediction using multiple samples.

Table of Contents

In this lab, we will review how to make a prediction for Linear Regression with Multiple Output.

• Build Custom Modules


Estimated Time Needed: 15 min

Class Linear

from torch import nn
import torch

Set the random seed:

torch.manual_seed(1)

<torch._C.Generator at 0x7f4538043150>

Set the random seed:

class linear_regression(nn.Module):
    def __init__(self,input_size,output_size):
        super(linear_regression,self).__init__()
        self.linear=nn.Linear(input_size,output_size)
    def forward(self,x):
        yhat=self.linear(x)
        return yhat

create a linear regression object, as our input and output will be two we set the parameters accordingly

model=linear_regression(1,10)
model(torch.tensor([1.0]))

tensor([ 0.7926, -0.3920,  0.1714,  0.0797, -1.0143,  0.5097, -0.0608,  0.5047,
         1.0132,  0.1887], grad_fn=<ViewBackward0>)

we can use the diagram to represent the model or object

we can see the parameters

list(model.parameters())

[Parameter containing:
 tensor([[ 0.5153],
         [-0.4414],
         [-0.1939],
         [ 0.4694],
         [-0.9414],
         [ 0.5997],
         [-0.2057],
         [ 0.5087],
         [ 0.1390],
         [-0.1224]], requires_grad=True),
 Parameter containing:
 tensor([ 0.2774,  0.0493,  0.3652, -0.3897, -0.0729, -0.0900,  0.1449, -0.0040,
          0.8742,  0.3112], requires_grad=True)]

we can create a tensor with two rows representing one sample of data

x=torch.tensor([[1.0]])

we can make a prediction

yhat=model(x)
yhat

tensor([[ 0.7926, -0.3920,  0.1714,  0.0797, -1.0143,  0.5097, -0.0608,  0.5047,
          1.0132,  0.1887]], grad_fn=<AddmmBackward0>)

each row in the following tensor represents a different sample

X=torch.tensor([[1.0],[1.0],[3.0]])

we can make a prediction using multiple samples

Yhat=model(X)
Yhat

tensor([[ 0.7926, -0.3920,  0.1714,  0.0797, -1.0143,  0.5097, -0.0608,  0.5047,
          1.0132,  0.1887],
        [ 0.7926, -0.3920,  0.1714,  0.0797, -1.0143,  0.5097, -0.0608,  0.5047,
          1.0132,  0.1887],
        [ 1.8232, -1.2748, -0.2164,  1.0184, -2.8972,  1.7091, -0.4722,  1.5222,
          1.2912, -0.0561]], grad_fn=<AddmmBackward0>)

the following figure represents the operation, where the red and blue represents the different parameters, and the different shades of green represent different samples.

About the Author:

Hi, My name is Juma Shafara. Am a Data Scientist and Instructor at DATAIDEA. I have taught hundreds of peope Programming, Data Analysis and Machine Learning.

I also enjoy developing innovative algorithms and models that can drive insights and value.

I regularly share some content that I find useful throughout my learning/teaching journey to simplify concepts in Machine Learning, Mathematics, Programming, and related topics on my website jumashafara.dataidea.org.

Besides these technical stuff, I enjoy watching soccer, movies and reading mystery books.

What's on your mind? Put it in the comments!