Linear regression is commonly used to quantify the relationship between two or more variables. It is also used to adjust for confounding. This course, part ofourProfessional Certificate Program in Data Science, covers how to implement linear regression and adjust for confounding in practice using R.
In data science applications, it is very common to be interested in the relationship between two or more variables. The motivating case study we examine in this course relates to the data-driven approach used to construct baseball teams described in Moneyball. We will try to determine which measured outcomes best predict baseball runs by using linear regression.
We will also examine confounding, where extraneous variables affect the relationship between two or more other variables, leading to spurious associations. Linear regression is a powerful technique for removing confounders, but it is not a magical process. It is essential to understand when it is appropriate to use, and this course will teach you when to apply this technique.
Start your review of Data Science: Linear Regression
Dray Presctt is taking this course right now.
This is supposed to be an introductory course in linear regressions using the programming language R. The problem is, there's no evidence that the professor knows how to program in R. None of his examples run. The course forums are full of people trying...
This is supposed to be an introductory course in linear regressions using the programming language R. The problem is, there's no evidence that the professor knows how to program in R. None of his examples run. The course forums are full of people trying to figure out how to get the code to work. Some of the problems seem to stem from the professor using his own private R packages, which define commonly-used R commands differently than in 'standard' R (neither the syllabus nor the lectures give any hint as to what packages he's using). Even if you download the packages that students have guessed, there are still some examples that nobody has figured out how to run, and these threads have been going on for months.
The professor seems to misunderstand basic programming concepts as well as syntax. A task was to find the oldest male child in every family, the professor's approach was to limit the data to entries where the child was the oldest and the child was male; in other words, in a family with a daughter and then a son, the professor's code would omit that family entirely.
Given the problems answering any of the questions or following any examples, it's tough to evaluate the professor's statistical knowledge.
Luiz Cunha completed this course, spending 3 hours a week on it and found the course difficulty to be easy.
A good course on Linear Models using R.
You get a lot of knowledge for the time spent.
Videos are concise, to the point.
Videos recaps with Key Points listing and videos examples codes are extremely useful.
Assessments are a clear application of the lesson concepts, without bugs and at the right complexity level.
The only minus is maybe the lack of depth, and low level of difficulty: I wish the course would have developped more on T-stats, F-stats, inconvenience of adding too many inputs to the linear model...
This Data Science Linear Regression course is absolutely a disappointment. In my opinion, it is not worth paying for this course. Since EdX is one of the most expensive online courses, I would expect high quality content. But this course is anything but high quality. Also, EdX's support is just arrogant and it takes forever to get an answer to a simple question. Mr. Irizarry is the worst teacher I’ve ever seen. He is not able to explain the content in an understandable way. There are a lot of free courses available on the internet. Use one of them instead.