CSCI 307
Data Mining
College of the Holy Cross, Fall 2023

Home | | Schedule | | Resources Python resoucres Anaconda is a great distribution to get started with Python for data mining applications, as it comes pre-installed with a number of useful libraries and has a good library management systems. For people more familiar with C++/Java, these can be useful. Introduction to Python for C++ users (Archived lecture notes from a past ICL course) Introduction to Python for Java users A video on Python strategies I liked The first library we will start studying is NumPy. And the official documentation is quite good. Beginner's guide Quickstart guide Beyond that, I quite like this chapter from the "Data Analysis with Python" book by Wes McKinney. Numpy Basics Next we start on Pandas. Which is great for reading in structured tabular data, such as in .csv files or such. It also has a lot of functionalities for data analysis and works seamlessly with numpy, matplotlib (which we'll see later) etc. There is a 10-minute intro in the Pandas official documentation, but it's somewhat bit messy. You might find this book to be a better reference as you keep doing more stuff on Python. The previously mentioned "Data Analysis with Python" book also has a great Pandas chapter The primary plotting library we will be using is matplotlib.pyplot. There's a very short tutorial here, that gives you lots of examples. Feel free to find the scatter plots in the page and compare it to what you do in HW2. PyPlot tutorial Math Resources Matrix and Linear Algebra The Essence of Linear Algebra series of videos from 3Blue1Brown is an excellent resource for gaining intuition about matrices, matrix operations and linear algebra. Principal Component Analysis StackExchange, StackOverflow, etc. are often great sources for learning. For example, this answer Explaining PCA is excellent.