Machine learning 2 - Neural Networks
1 like527 views
Neural networks are machine learning models inspired by the human brain. The document discusses neural network architectures like perceptrons and multi-layered deep neural networks. It also covers common neural network techniques for images like convolution and pooling layers used in convolutional neural networks (CNNs). Finally, it provides code examples for using neural networks in Python with the Keras library on image datasets like MNIST and examples of loading pre-trained models like ResNet50 and VGG16 to classify images.
Machine learning 2 - Neural Networks
- 1. Machine Learning Part 2. Neural networks
- 3. Neuron
- 4. Perceptron
- 7. Problems with images Too big vectors (200x200x3 = 120,000) Pixel position matters
- 8. Convolution
- 10. CNN
- 12. pip install keras open/create ~/.keras/keras.json edit "image_dim_ordering": “th”; “backend”: “theano”; open script from: http://bit.ly/ empatika-keras-1 MNIST
- 13. Dogs & Cats cd <folder that contains ‘data’ subfolder from Google Drive> open/create ~/.keras/keras.json edit “backend”: “tensorflow”; open script from: http://bit.ly/ empatika-keras-2 (initial gist)
- 14. Pre-trained models ResNet50 VGG16 cd <to deep-learning-models subfolder>