sukhjitsehra / machinelearning

supervised and unsupervised algorithms from Andrew Ng's machine learning class

Machine Learning

The majority of the material here was created while taking Andrew Ng's free online Machine Learning class which I highly recommend!

"A computer program is said to learn from experience E with respect to some task T and some performance measure P, if its performance on T, as measured by P, improves with experience E."

~ Definition of Machine Learning by Tom Mitchell

How to use this code

1. Install [Octave free here](https://db.tt/J97Im052) or [Matlab not free here](http://www.mathworks.com/products/matlab/). Note that Octave = Matlab without the nice graphical user interface. I use Octave so don't feel like you are missing anything if you don't have money for Matlab.

2. Fork this repository and clone it locally! Navigate into specific folders (made them very specific) and look at the README.md file for that specific folder for which file(s) to run to see examples of what machine learning algorithms can do for you. Enjoy!

What each file/folder in this repository is for:

- [diagnosticTests](./diagnosticTests) = tests that will give you insight into what is & isn't working with a learning algorithm

• imagesForExplanation = contains images used in other folder's README.md files for explanation so don't worry about this folder

• supervisedLearning = teach the computer how to learn

  • linearRegressionIn1Variable
  • linearRegressionInMultipleVariables
  • logisticRegression
  • LinearAlgebraReview.md

• unsupervisedLearning = let the computer learn by itself

  • neuralNetworks
    • digitRecognition
    • learningWithBackpropagation

• README.md = the file you are reading right now