Build an Artificial Neural Network (ANN) using PyTorch - Example with 'USPS' dataset
In this video, I tried to perform image classification by implementing Artificial Neural Network (ANN). I used the USPS dataset, which contains handwritten digit images of 16 x 16 pixels. ANNs are composed of artificial neurons. Each neuron has inputs and produces a single output. The outputs of the final neurons accomplish tasks such as image classification.
GitHub address: https://github.com/randomaccess2023/M...
For more details: https://en.wikipedia.org/wiki/Artific...
Starting point of each section:
01:28 --- Import the libraries
03:04 --- Define the device
05:09 --- Load USPS dataset
07:02 --- Define dataloaders
08:12 --- Define the network architecture
11:12 --- Visualize the model architecture
12:14 --- Define the train batch
14:02 --- Define the test batch
15:04 --- Define the accuracy function
17:02 --- Set up model (ANN), criterion (loss function) and optimizer
17:57 --- Train the model
26:43 --- Plot model loss and accuracy
30:17 --- Model prediction
#artificial_neural_network #supervised_learning #image_classification #usps_dataset #data_science #jupyter_notebook #python #pytorch #ann