This is an intermediate-level specialization, and it is assumed you already know how to write programs in Python. It also assumes some elementary knowledge of statistics and discrete mathematics, but nothing too advanced.
As hinted at by the word “Applied” in the specialization and course titles, there is not much theory presented in this course. There is just enough theory to understand the exercises.
Suggestion: Take the specialization concurrently with Andrew Ng’s deeplearning.ai specialization, so you will get a nice dose of neural-network theory mixed in with data science, but people who want to understand the algorithms in detail will need further study.
Specification: Michigan University Applied Data Science with Python Specialization
|Free Trial (in days)|
|Duration (in months)|
|What will you learn?|
|Programming Language Used|
1. Introduction to Data Science in Python – learn basics of the python programming environment. Students will be able to take tabular data, clean it, manipulate it, and run basic inferential statistical analyses.
2. Applied Plotting, Charting & Data Representation in Python – learn visualization basics, with a focus on reporting and charting using the matplotlib library.
3. Applied Machine Learning in Python – identify the difference between a supervised (classification) and unsupervised (clustering) technique, identify which technique they need to apply for a particular dataset and need, engineer features to meet that need, and write python code to carry out an analysis.
4. Applied Text Mining in Python – text mining and text manipulation basics.
5. Applied Social Network Analysis in Python – learn network analysis through tutorials using the NetworkX library
|Graded quizzes & assignments|
2 reviews for Michigan University Applied Data Science with Python Specialization
- Solid introduction to what libraries are available and how to use them, with enough challenge to make you dig into the documentation and Stack Overflow a lot.
- Most of the courses include some lectures or assignments dealing with the ethics of data science.
- The exercises are automatically graded, and the auto-grader has a lot of problems.
- Assignments that should take an hour or two instead take two or three times that due to the auto-grader rejecting correct answers with a "wrong" data type, or "wrong" number of significant figures, or expecting a different result than what the assignment's instructions specify. After doing each assignment, one has to spend an hour reading the course forums to find out the tricks to getting correct work accepted.