These projects illustrate some basic concepts in Deep Learning.
- Perceptron.ipynb: Implement a NN that performs binary classification. Use the model to predict the classification of a single point. Uses contour diagrams to visualize the output.
- Logistic_Regression.ipynb: Use logistic regression to implement binary classification.
- logistic_regression_animated.py: Use logistic regression to implement binary classification. Run this from the command line (outside of PyCharm or other IDE) to see the 'animation' of the line as the gradient descent function does its work to find the best fitting line.
- Polynomial_Regression.ipynb: When working with non-linear data, this shows how to implement polynomial regression on a set of data points.
- Deep_Neural_Network.ipynb: Use Keras to implement a DNN that classifies data that is not linearly separable.
- Multiclass_Classification.ipynb: Use Keras to build and train an neural network that can classify more than two types of data.
- MNIST_Image_Classification.ipynb: Use Keras to build a model that classifies digits from the MNIST dataset.
- CNN_MNIST.ipynb: Implement a CNN using a LeNet-based architecture that classifies digits from the MNIST dataset.