Here deep learning algorithm is applied on CIFAR10 dataset for image classification. CNN parameters are tunned with different values for finding the best parameter values.
Dataset consists of 60000 colour (r * g * b) images of size 32x32 with 10 classes.
- Airplane
- Automobile
- Bird
- Cat
- Deer
- Dog
- Frog
- Horse
- Ship
- Truck
- Tensorflow
- Matplotlib
- Numpy
- relu
- softmax
- sigmoid
- adam
- nadam
Here the CNN alogorithm evaluated with different filter values, activation functions and optimizers. The intial CNN with specified parametres are effective compared to other codes executed over here.