I am looking into the Data Science program at Eastern. I have been moving progressively upward in my organization as a “data specialist”. I have no related formal training and thought this could be a good option for continued growth. Math has always been a weak point. I use formulas in Excel to do my math for me. I was told that the program at Eastern uses algebraic statistics. I am not entirely sure what that entails. Does anyone know places I could try my hand at it? I don’t want to go too far into algebra beyond what I might actually need.
Algebra-based is chosen when a program doesn't want to require high math prerequisites. It appears that Eastern University has no math prerequisites so they can accept almost everyone without the barrier of a calculus prerequisite.
This is HEAVY stuff. Every time I see the word "combinatorics" I shudder. I've been told by GOOD math students (which I'm not) that it's pretty deadly. "Combinatorics and optimization" is the usual lethal combo. And I see combinatorics here. They call this an "Intro" to Algebraic Statistics. Some intro! See for yourself... https://www3.diism.unisi.it/~chiantini/did/00Book.pdf There are some videos out there on Algebraic Statistics. Maybe watch a couple of those. For those who like wine, some of your favourite might help. Just a suggestion.
I have never heard the word “Combinatorics” in my entire life. Yeah, this is why I might need to try out the applicable maths first.
Cool. Retro. Comfortable. Reminds me of the "number" songs on Sesame Street, when my kids (now in their 50s) were small. Thanks, Jonathan.
I think, around where I live, it's mostly about a 3rd year word in a 4 year math degree. I remember, a very bright friend of my son's, many years ago, telling me how much trouble it was giving him in University. He had to re-take the course - the only time that ever happened to him. He recovered, earned his degree and worked for many years, 'way up in the higher echelons of banking. I think he's recently retired, now.
With me, everything was a blur after long division. And I left England at age 9, able to do everything with pounds, shillings pence, florins. farthings and threepenny bits - but hadn't seen decimals yet. OK. I adapted. A bit later I got to Algebra. I used language, not math, skills to "translate" word problems into equations. That worked OK but then I couldn't SOLVE the equations. Yeowch!! Once I grew up, everything was fine. Somehow all the stuff "jelled" after I left school. 25 years later, I took three semesters of business and financial math in College, got straight A's and actually used it in University courses. I absorbed a bit more math in other college courses - e.g. some basic trigonometry in house-building. I could actually cut rafters pretty well! I'm no longer afraid of math now -- except maybe for combinatorics --- yikes! But that's conditioning...it can be overcome.
whoops - that's Advanced - not algebraic. Bad Google! Udemy? This one has combinatorics in it... https://www.udemy.com/course/math-for-data-science-masterclass/ I looked. Lots of Uni-based courses. Not much else. Some 50-minute Intro videos and stuff. Maybe you want to look at those for a start. here's one. This is 9 minutes by Seth Sullivan who wrote a textbook on Algebraic Statistics. An hour intro, By a UC Berkeley professor.
You will not need any math for the Eastern program. The program covers some of it in theory but the only math you will need to do in practice is a teensy bit of matrix math. This doesn't require any pre-education. Brushing up on linear algebra and calculus can be useful to understand the theory behind the algorithms but is not necessary to do the assignments. Edit: Forgot the stats part. Algebraic statistics are non-calculus statistics. This is the basic stuff like mean, median, mode, hypothesis testing and similar. Fairly basic stuff.
Thanks, Dustin! When I Googled it - they came up with stuff like this: "Algebraic statistics is concerned with the development of techniques in algebraic geometry, commutative algebra, and combinatorics, to address problems in statistics and its applications." And this too - on the same Preface page: "Algebraic statistics is a relatively new field (emphasis mine - J.) that has developed and changed rather rapidly over the last fifteen years. One of the first pieces of work in this area was the paper of Diaconis and the second author [33], which introduced the notion of a Markov basis for log-linear statistical models and showed its connection to commutative algebra. From there, the algebra/statistics connection spread to a number of different areas including the design of experiments (highlighted in the monograph [74]), graphical models, phylogenetic invariants, parametric inference, algebraic tools for maximum likelihood estimation, and disclosure limitation, to name just a few." From here: https://math.berkeley.edu/~bernd/owl.pdf (from the preface page.) The UC Berkeley Professor makes it sound like "New Wave Rocket Science" and you bring the rocket safely back down to Earth. Pure genius. Thanks.