Machine Learning Foundations

In this repository I collect all the completed assignements for the course "Machine Learning" given by Andrew NG through Coursera. All the functions have been written in Matlab by myself. I followed the instructions given for each exercise and in the learning sections.

You can find here examples/functions about:

1. LINEAR REGRESSION

• plotData.m - Function to display the dataset
• computeCost.m - Function to compute the cost of linear regression
• gradientDescent.m - Function to run gradient descent
• computeCostMulti.m - Cost function for multiple variables
• gradientDescentMulti.m - Gradient descent for multiple variables
• featureNormalize.m - Function to normalize features
• normalEqn.m - Function to compute the normal equation

2. LOGISTIC REGRESSION

• plotData.m - Function to plot 2D classification data
• sigmoid.m - Sigmoid Function
• costFunction.m - Logistic Regression Cost Function
• predict.m - Logistic Regression Prediction Function
• costFunctionReg.m - Regularized Logistic Regression Cost

3. MULTICLASS LOGISTIC REGRESSION & NEURAL NETWORK

This exercise is an implementation of one-vs-all logistic regression and neural networks to recognize hand-written digits.

One VS all - multiclass classification

• displayData.m - display 2D a number of digits represented by 20x20 pixels in a grayscale
• lrCostFunction.m - Compute cost and gradient for logistic regression with regularization
• oneVSall.m trains multiple logistic regression classifiers and returns all the classifiers in a matrix (i.e. the trained parameters
  theta). fmincg.m is called for minimize the cost and find optimal theta.
• predictOneVsAll.m - predict the labels for a trained one-vs-all classifier

Neural network

(The parameters theta are already trained in this exercise)

• predict.m - predict the label of an input given a trained neural network

4. NEURAL NETWORK LEARNING

• displayData.m - Function to help visualize the dataset
• fmincg.m - Function minimization routine (similar to fminunc)
• sigmoid.m - Sigmoid function
• computeNumericalGradient.m - Numerically compute gradients
• checkNNGradients.m - Function to help check your gradients
• debugInitializeWeights.m - Function for initializing weights
• predict.m - Neural network prediction function
• sigmoidGradient.m - Compute the gradient of the sigmoid function
• randInitializeWeights.m - Randomly initialize weights
• nnCostFunction.m - Neural network cost function

This exercise is an implementation of a neural network model to recognize hand-written digits and goes through the whole process of training a nn:

• Pick a n architecture
• Random initialization of parameters/weights
• Implementation of the Forward Propagation
• Implementation of the cost function
• Implementation of the Back Propagation
• Gradient Checking
• Minimization of the cost function
• Prediction and accuracy

4. REGULARIZED LINEAR REGRESSION

• linearRegCostFunction.m - Regularized linear regression cost function
• trainLinearReg.m - Trains linear regression using your cost function to estimate theta
• polyFeatures.m - Maps data into polynomial feature space (from linear to polynomial model)
• featureNormalize.m - Feature normalization function
• fmincg.m - Function minimization routine (similar to fminunc)
• plotFit.m - Plot a polynomial fit
• learningCurve.m - Generates a learning curve(Jtrain and Jcv)
• validationCurve.m - Generates a cross validation curve

In this exercise, one implements regularized linear regression to predict the amount of water flowing out of a dam using the change of water level in a reservoir. Then one goes through some diagnostics of debugging learning algorithms and examine the effects of bias v.s. variance.