ML-Workshop-for-THW

Doing some fun machine learning with "The Hacker Within" community, on 3rd October

As Scientists you may be familiar with Machine Learning before. If you have played with some sort of function fitting then you have seen some aspect of Machine Learning(ML).

It is the current HOT topic, and it looks good on your resume. People think its alchemy and some people have very strong opinions about it. But I am going to show you how you can use it in your daily life (or scientific career).

MACHINE LEARNING, WHAT IS?

Machine learning is a glorified way of finding a function iteratively. (Basically, Newton's method, on steroids). These functions need not be something analytic nor should look like any standard function you think of when you hear the word function.

ML is useful for forecasting, finding patterns, function fitting, classifying and plenty more.

Machine Learning usually goes in the following steps:

1. What is your problem?
2. Getting and preparing your data
3. Try out different algorithms (or think of a few suitable ones)
4. Evaluate your results
5. Improve your results
6. Start all over again (usually from 3)
7. ???
8. Profit

Today, we will go through the above steps (Till 4) and solve a problem.

ALL ABOARD THE TITANIC

Let's see if you can survive the titanic. Download the data set from my github, or https://www.kaggle.com/c/titanic/data

The whole tutorial is in the ipython notebook. Install Jupyter and open that! (THW Machine learning tutorial.ipynb)