Skip to content

Latest commit

 

History

History
28 lines (19 loc) · 2.35 KB

README.md

File metadata and controls

28 lines (19 loc) · 2.35 KB
Raw

Machine Learning Programming Exercises

  • Programming Exercise 1 - Linear Regression In this exercise we will implement Linear Regression and get to see how it works on data. Linear Regression parameters (theta) are fit using Gradient descent algorithm. We also plot initial data and the data with linear regression fit. The plot of Cost function and Contour plots are also part of this exercise.

  • Programming Exercise 2 - Logistic Regression In this exercise we will implement Logistic Regression and apply it to two different datasets. We also get introduced to Sigmoid function. Cost function and Gradients are calculated without and with Regularization and decision boundaries are plotted.

  • Programming Exercise 3 - Multi-class Classification and Neural Networks In this exercise we will implement One-vs-All Logistic Regression and Hand-written digit recognition using Neural Networks.

  • Programming Exercise 4 - Neural Networks Learning In this exercise we will implement Back propagation algorithm for Neural Network training and apply it to the task of Hand-written digit recognition.

  • Programming Exercise 5 - Regularized Linear Regression and Bias vs Variance In this exercise we implement regularized Linear Regression and we use it to study models with different Bias-Variance properties.

  • Programming Exercise 6 - Support Vector Machines (SVM) In this exercise we will implement a Support Vector Machine to build a Spam classifier.

  • Programming Exercise 7 - K-means clustering and Principal Component Analysis (PCA) In this exercise we will implement K-means clustering and apply it to compress an image. In the second part we use Principal Component Analysis to find a low-dimensional representation of face images.

  • Programming Exercise 8 - Anamoly Detection and Recommender Systems In this exercise we will implement Anamoly detection and apply it to detect failing servers on a Network. In the second part we use Collabarative filetering to build a recommender system for Movies.

  • Credit Card Fraud Analysis - In this exercise we analyse real Credit Card transaction data and classify whether a given transaction is fradulent or valid using Logistic Regression. The dataset used is real but it is not tagged for confidential reasons. Performance metrics like Precision, Recall and F1 Score are computed as this a highly skewed dataset. Please check the source for more details.