Scikit-learn Tutorial: Machine Learning in Python

Scikit-learn is a free machine learning library for Python. It features various algorithms like support vector machine, random forests, and k-neighbours, and it also supports Python numerical and scientific libraries like NumPy and SciPy. In this tutorial we will learn how to easily apply Machine Learning with the help of the scikit-learn library, which was created to make doing machine learning in Python easier and more robust. To do this, we'll be using the Sales_Win_Loss data set from IBM's Watson repository. We will import the data set using pandas, explore the data using pandas methods like head(), tail(), dtypes(), and then try our hand at using plotting techniques from Seaborn to visualize our data. Then we'll dive into scikit-learn and use preprocessing.LabelEncoder() in scikit-learn to process the data, and train_test_split() to split the data set into test and train samples. We will also use a cheat sheet to help us decide which algorithms to use for the data set. Finally we will use three different algorithms (Naive-Bayes, LinearSVC, K-Neighbors Classifier) to make predictions and compare their performance using methods like accuracy_score() provided by the scikit-learn library. We will also visualize the performance score of different models using scikit-learn and Yellowbrick visualization. If you need to brush up on these topics, check out these pandas and data visualization blog posts. For this tutorial, we will use the Sales-Win-Loss data set available on the IBM Watson website.