-Load two common datasets
-Use any needed pre-processing function to analyze dataset
-Use LeNet-5 to complete the process of classification
-Print LeNet-5 architecture
-Print number of Trainable parameters in each layer
-Print confusion matrix relative to testing samples
-Print precision, recall, f1_score
-Comment on your results
1- Import Libraries we will use
2- Load Datasets:
-gray scaled images dataet (MNIST).
-Colored images dataset (CFIAR-10).
3-Data Preproccesing :
-Reshape taining and testing data.
-Normalization of Training and Tetsing data.
4- Plot Figure for each Dataset.
5- Split the Training Dataset into train and Validation.
6- Build The LeNet-5 Architecture:
-2 CNN Layers each layer has differnce filter size.
- we used tanh acctivation function.
-Average Pooling.
-Flatten dataset.
-Build 2 NN Layers with different number of neurons.
-build the output NN Layer, we used softmax acctivation function.
7- Print The LeNet model
8- Model Training using Adam Optimizer and crossentropy as Loss Function.
9- iterate over layer to print the number of trainable parameters.
10- Print Confusion matrix relative to testing samples.
11- Calculate Precision, Recall, and f1_score.
