fferrin / Coursera-Machine-Learning

Machine Learning

Code for programming assignments in Octave from the Coursera course Machine Learning, given by Andrew Ng for Stanford University.

Assignment 1 - Linear Regression (100%)

• computeCost.m - Compute cost for linear regression.
• computeCostMulti.m - Compute cost for linear regression with multiple variables.
• featureNormalize.m - Normalizes the features in a vector.
• gradientDescent.m - Performs gradient descent to learn parameters.
• gradientDescentMulti.m - Performs gradient descent to learn parameters with multiple variables.
• normalEqn.m - Computes the closed-form solution to linear regression.
• warmUpExercise.m - Example function in octave.

Assignment 2 - Logistic Regression (100%)

• costFunction.m - Compute cost and gradient for logistic regression.
• costFunctionReg.m - Compute cost and gradient for logistic regression with regularization.
• predict.m - Predict whether the label is 0 or 1 using learned logistic.
• sigmoid.m - Compute sigmoid function.

Assignment 3 - Multiclass Classification and Neural Networks (100%)

• lrCostFunction.m - Compute cost and gradient for logistic regression with regularization.
• oneVsAll.m - Trains multiple logistic regression classifiers and returns.
• predict.m - Predict the label of an input given a trained neural network.
• predictOneVsAll.m - Predict the label for a trained one-vs-all classifier.

Assignment 4 - Neural Networks Learning (100%)

• nnCostFunction.m - Implements the neural network cost function for a two layer neural network which performs classification.
• sigmoidGradient.m - Returns the gradient of the sigmoid function.

Assignment 5 - Regularized Linear Regression and Bias vs. Variance (100%)

• learningCurve.m - Generates the train and cross validation set errors needed to plot a learning curve.
• linearRegCostFunction.m - Compute cost and gradient for regularized linear regression with multiple variables.
• polyFeatures.m - Maps X (1D vector) into the p-th power.
• validationCurve.m - Generate the train and validation errors needed to plot a validation curve that we can use to select lambda.

Assignment 6 - Support Vector Machines (100%)

• dataset3Params.m - Returns your choice of C and sigma for Part 3 of the exercise.
• emailFeatures.m - Takes in a word_indices vector and produces a feature vector from the word indices.
• gaussianKernel.m - Returns a radial basis function kernel between two points.
• processEmail.m - Preprocesses a the body of an email and returns a list of word indices.

Assignment 7 - K-means Clustering and Principal Component Analysis (100%)

• computeCentroids.m - Returs the new centroids by computing the means of the data points assigned to each centroid.
• findClosestCentroids.m - Computes the centroid memberships for every example.
• pca.m - Run principal component analysis on the dataset X.
• projectData.m - Computes the reduced data representation when projecting on eigenvectors.
• recoverData.m - Recovers an approximation of the original data when using the projected data.

Assignment 8 - Anomaly Detection and Recommender Systems (100%)

• cofiCostFunc.m - Collaborative filtering cost function.
• estimateGaussian.m - Estimates the parameters of a Gaussian distribution using the data in X.
• selectThreshold.m - Find the best threshold to use for selecting outliers.