I hope this will be useful to anyone interested in doing a year abroad of studies. I am going to give a personal overview of my experiences and some (hopefully) useful advice!
If you have any questions after reading this post, please feel free to reach out to me at
Living in Switzerland
Switzerland is a great place to live in, especially if you are a student.
- Visa: Apply as soon as you receive your letter of acceptance from EPFL at the Embassy of Switzerland in London. The Visa will be valid for a few months until you receive your resident permit.
- Travel: London to Geneva by plane (about 20 - 40 GBP in August 2017 and a 1-hour flight). Then Geneva to Lausanne by train (20-30 CHF and less than an hour journey. You could also buy it in advance on the SBB/CFF app, which is usually cheaper).
- Transport: You can purchase travel passes on TL (Transport Lausanne). You need to go to an office (there’s one at EPFL) to sort out your registration. This gives you access to buses and the metro in zones 11/12. Check the subscriptions on their website.
EPFL metro station:
Useful things to take with you
- Bank statements
- Student status statement (from Student Hub)
- Passport photos
- Acceptance letter from EFPL
- Motivation to work hard
- Deadline to register for exams is 2 weeks into the semester and the deadline to withdraw from an exam is around 1 or 2 months before the exams. So I recommend taking 7-9 courses per semester at the start and then slowly dropping to 6 or 7 when it comes to the exams, just in case you change your mind.
- Coursework is similar to Imperial, although the project report criteria are stricter.
- Expensive: around 10% more expensive than London. A meal is usually 20-40 francs at restaurants, although university canteens are usually 8-14 francs. Grocery store shopping is also slightly more costly, but at the same time the food quality is usually very good.
- Accommodation at FMEL is guaranteed, so only 510 CHF / month! Very big rooms and nice flat.
- Understand how recycling and rubbish collection works!
- Realise that there is a green box for compost
- Keep quiet at night
- Keep the kitchen and toilets very clean. There are regular checks from the house manager
- Do not lose your key or else it will be an expensive lesson
- Buy duvet/blankets, pillows and bedsheets from Amazon
- Buy your mugs, plates etc…
Student hall that I lived in!
- Recommended restaurants: Restaurant Henan (near Renens), Chalet Suisse (on top of the hill in Lausanne), El Chiringuito (near Lausanne-Flon), Café Romand (near Flon)
- EPFL canteens: Thai food truck (good food ;) ), Le Vinci, Le Parmentier (oh ffs the same food again as lunch…), Thai canteen next to Le Parmentier
This is actually made of crack!?
宫保鸡丁 from Henan though hmmm:
Local places to visit
- Geneva - CERN
- Morge – La fête de tulip and Audrey Hepburn’s residence
- Fribourg – The streets and innovation house (winner of a global smart houses competition)
- Gruyère – Cheese and chocolate
- Vevey - Wineyards
- Zurich – ETH Zurich and FIFA
Life at EPFL
- EPFL is a really nice place if you want to do research or work on personal projects. There are academic seminars every week in the maths department – Topology seminar, Numerical analysis, Analysis, Statistics. For example, Kathryn Hess works on topological data analysis and topological genetics.
- The coding club does programming competitions during the year (they only host a few events sadly)
- Hyperloop is pretty big (3rd in the Elon Musk Hyperloop competition this year). If you want to join you can apply in September.
- ESN hosts fun events every week for exchange students. They also have Chalet weekends at the beginning of semesters (not to be missed!).
- There is also a Go society which looks very fun
- The consulting society is also very active, and hosts workshops every week
- The badminton club has recreational sessions every week on Tuesday nights
Annual dinner organised by CQFD (MathSoc of EPFL). Guess which team I was on
- Measure and integration
- Introduction to PDEs
- Functional analysis
- Advanced numerical analysis
- Linear models
- Deep Learning
- Applied Biostatistics
- Modern Regression Methods (highly recommend! Taught by Davison, who did his PhD at Imperial)
- Bayesian computation
- Applied stochastic processes
- Martingales and applications
- Time series
- Anthony Davison
- Michael Kapralov (Theory of Computation)
- Thomas Mountford
- John Maddocks
- Martin Odersky (CS, founder of Scala)
Interesting courses that I did not take:
- Signal processing (it’s on Coursera too!)
- Probability theory
- Introduction to differentiable varieties
- Risks, rares and extremes (Davison)
- Robust and non-parametric statistics (Mogenthaler)
- Introduction to numerical PDEs (Fabio Nobile). Basically a finite elements course
- Quantum Calculus
The courses are not as well-moderated as Imperial courses. There are potential pitfalls, for example poor teaching, poor exams, poor content etc…
A solution is to take more courses at the beginning. This way you can avoid doing something you don’t enjoy.
As aforementioned, there are some very good courses. Notably taught by Anthony Davison and Michael Kapralov.
That boardwork by Mountford though…
- Speak to someone who has done an exchange or studied at EPFL before
- Work very, very hard during your first few months
- Take advantage of the connections that you can build with friends, classmates and professors
Beautiful views in Renens:
Going off on a tangent
So here are some really good courses you should think about at EPFL. I’ve organised them in categories of specialisation.
- Statistics and Probability
- Risk, rares and extremes (MATH)
- Mathematics of psychology (MATH)
- Modern regression methods (MATH)
- Linear models (MATH)
- Applied stochastic processes (MATH)
- Maths of psychology (or something like that. It’s in MATH)
- Bayesian computation (MATH)
- Time series
- Robust and non-parametric statistics (MATH). Taught by the expert Stephan Morganthaler.
- Biostatistics (MATH)
- Mathematics of DNA (not really a stats or probability course but I’ll put it here)
- Data science, Machine Learning, Deep Learning
- Deep learning (EE) or Artificial neural networks (CS)
- Introduction to databases. It’s a second year course and teaches you database design and SQL
- Machine learning (CS). Teaches you the basics of machine learning and deep learning (with TensorFlow).
- Information theory and signal processing (useful if you want to do computer vision, word embeddings etc…)
- Introduction to natural language processing. Useful if you want to work on word embeddings.
- Software engineering (CS). Gives you a nice coding background. I think you code in Java (I learned Java in 1 week. Trust me, it’s not that hard if you know Python and R)
- Functional programming (CS). This is a second year course that teaches you functional programming with Scala (the guy who invented it teaches at EPFL!). If you want do parallel computing (stuff like Hadoop, Spark and MapReduce) in the future as a machine learning guru, then take this course.
- Internet analytics (CS). Applied course with Python. Good Python practice.
- Statistical machine learning (MATH). Teaches you rigorous machine learning.
- Algebra and Geometry
- Introduction to differentiable varieties (MATH). Gives you a good background in differential geometry.
- Homology and homotopy (MATH). I’m not an expert but if you are a data scientist, learning this will allow you do do stuff like persistent homologies and simplices in topological data analysis.
- Differential geometry of curves and surfaces. Taught by Maddocks
- Lie algebras
- Analysis, PDEs and CPDEs
- Functional analyses I and II (MATH)
- Numerical approximation of PDEs I (MATH), not II. It’s taught by Fabio Nobile who is an excellent lecturer
- Analysis on groups (MATH)
- Harmonic analysis (MATH). Taught by Krieger. Seems pretty tough and the exam is oral, which makes it even harder. Maybe taking this course but not the exam would be better. Very Princeton-esque
- Probability theory (MATH). Taught by Mountford. Be very careful. If you get a 5.5+, you are pretty much guaranteed a PhD here.
- Computational linear algebra (MATH)
- Computational finance (MATH)
- Numerical stochastic differential equations (MATH). Good teacher, Assyr.
- Stochastic calculus (MATH). Taught by Mountford.
Any course taught by Davison (sampling methods and bootstrap), Krieger (PDEs), Kapralov (pure computer science) and Mountford (crazy inequalities in probability theory) are worth it.