Long story short, I'm really interested in machine learning research and aim to get into a top CS PhD program. I recently got admitted to the best CS program (5 year bsc+msc) in my country but it is not exactly MIT. My main concerns are the lack of math and reputation. I am therefore considering starting another bachelor's in mathematical statistics in parallel.

Assuming I have the talent, ambition and drive to do it with good GPA and side projects, should I?

How beneficial would this be from a competence stand point for ML research?

How beneficial would this be for CS PhD admissions?

Math in CS program:

Single variable calculus

Multi variable calculus

Algebra and geometry (Linear algebra without proofs)

Discreet mathematics

Modelling and simulation

Numerical methods

Probability and statistics

Logic for computer scientists

Math in Bsc. Mathematical statistics:

Single variable calculus

Multi variable calculus

Linear algebra (with proofs)

Linear algebra advanced course

Discreet mathematics

Algebra and combinatorics

Analysis A

Analysis B

Foundations of analysis advanced course

Complex analysis

Abstract Algebra

Mathematical logic

Probability 1

Probability 2

Statistical analysis

Stochastic processes and simulation 1

Stochastic processes and simulation 2

Categorical data analysis

Linear statistical models

Theory of statistical inference

So what do you all think?


A PhD is a first and foremost a research degree. Having some research experience will be much better than a whole bunch more bachelor's level classes. Instead of doing a double bachelor's, spend that time doing research during your MS degree. Having a published journal or conference paper will probably go a lot further in an application than a double BS

| improve this answer | |
  • What would be better; A double bachelor's with some research experience or a single bachelor's with an exceptional track record of research and side projects related to machine learning? – KTH Aug 10 '19 at 17:15
  • In my opinion, the later is far far better (assuming that "exceptional track record" includes published papers) – Daniel K Aug 10 '19 at 17:52

Not the answer you're looking for? Browse other questions tagged or ask your own question.