Teaching

Universite de Franche-Comte (2013-2014)

As a PhD student at the Universite de Franche-Comte in France, I had the chance to assist in teaching a variety of undergraduate and lower graduate courses. This involved class lecturing in addition to laboratory experiments planning, supervision, and reports reviewing, as summarized below.

• Lab, Continuum mechanics (32h) for 2nd year (Geo-sciences), and 3rd year (Physics).
This is an introductory course to classical mechanics and thermodynamics. The experiments included for example applications to the Bernoulli theorem, thermo-couples, Pitot tubes, and measuring moments of inertia. It is also the first lab course for many of the second year students. I hence had to start by explaining the fundamentals of physical experiments such as the different types of measurement uncertainties, the ins and outs of scientific data plotting, and how to write a university level report. While the science I had to convey was simple, this course has its own challenges. As a start it was my first serious attempt of teaching at a university level. More importantly however, I was 23, barely a couple years older than some of the students. Drawing professional lines was therefore both challenging and crucial.

• Lab, Numerical treatment of physical measurements (24h) for 2nd year (Chemistry).
This was a computer lab course focusing on the statistical treatment of physical measure- ments. I had to explain to students some fundamental notions of statistics, such as the dif- ferent statistical distributions (Gaussian, Poisson, Chi-squared, etc), and fitting these to data using Excel. The main (minor) challenge I faced teaching this course was my ignorance of Excel, having used more science-grade tools throughout my career. I hence had to make sure that I am extra well prepared before every lecture.

• Lectures and tutorials, Planetary Sciences for Master students (8h).
This is a module of a graduate level course that my PhD advisor was supposed to teach initially, but then delegated to me. Scientifically, it was the most exciting course I had taught to date as it dealt with subjects very close to my own research: an introduction to solar system sciences and planet formation. I hence started by explaining the basics of stellar formation (that can be understood intuitively using the Virial theorem), then the physics of proto-planetary disks, and finally planet growth through the accretion of solids and gas. The main challenge in teaching planet formation theory is how little we know as a fact. Many aspects of our current knowledge on the subject are conjectural at best. This is fundamentally due to the intrinsic difficulty in observing the process of planet formation directly. It is hence crucial to emphasis to students that while our understanding of some aspects (how stars form, etc..) is robust, other facets are still pretty much in flux among scientists

University of Toronto (2016-2018)

As a postdoc at the University of Toronto, my main job consisted mostly of research in planetary sciences. I did nonetheless contribute to local teaching activities.

• Machine Learning open workshop (35h).
My first teaching experience at UofT was in 2016 when, with fellow postdocs, decided to organize and run a 5 days long workshop on Machine Learning and its applications in plane- tary sciences (Materials available here). The workshop consisted of theoretical introduction to the different Supervised and Unsupervised machine learning techniques, with applications to astronomy. This was a hands on workshop, with all material delivered in the form of fully interactive “Jupyter Notebooks” (example shown in Fig. 1). The large and diverse audience consisted of undergraduate and graduate students from most scientific majors, in addition to interested faculty members. Creating well balanced lectures of interest to everyone was hence particularly challenging.

• Lectures and Lab “Introduction to Machine Learning for Industry and Science” (30h).
The success of our machine learning workshop led to the creation of an official, new, upper undergrad course on the same topics (EESC24). I had the opportunity of both designing the syllabus (proposal attached here) and then teacher-assisting the professor (Diana Valencia). The main challenge teaching this course was simply the heavy workload, where I had to correct/grade the weekly assignments that consisted of solving machine learning problems using computer codes.

American University of Beirut (2018)

In 2018 I was invite to spend one semester as a visiting assistant professor at the American Uni- versity of Beirut. AUB is a private, selective, and fully English speaking university in Beirut, Lebanon. It is the highest ranking educational institution in the Arab world. I thought it would be an interesting experience and thus agreed. There I taught two courses, organized a workshop, and interacted with many students.

• Lectures “PHYS210: Introductory physics for engineers” (36h).
This is the main fundamental physics course all engineering students have to take. It is a thorough course covering a wide range of classical physics topics such as Newtonian me- chanics, thermodynamics, fluids, and optics. The detailed syllabus is attached here. As the main professor, I had to prepare the presentations and lecture to 120 students in the hall. I was also involved in preparing homework, exams, and coordinating with the TAs and lab managers. The huge responsibility I felt from being completely in charge, and the large number of students in class, made this my most challenging teaching experience to date.

• Lectures “Introduction to planetary sciences” (36h).
This is an undergraduate elective I fully designed and taught myself. It is an umbrella course introducing students to a wide variety of planetary science topics ranging from Keplerian mechanics to atmospheres, surfaces, and planet formation. The full syllabus is attached here. This truly interdisciplinary course interested students majoring in physics, chemistry, geology, applied math, and engineering. My philosophy when designing it was to introduce the largest number possible of students to the field, as I was only a visiting Prof. and thus the course would be available for only one semester. The main challenge was hence how to create a memorable course that is interesting to a diverse body of students with different backgrounds and levels of on-topic knowledge. One month after the start of the course I conducted an anonymous online evaluation survey, with the (very positive) results shown in Fig. 2.

• Introduction to scientific computing with python workshop (2 full days).
The idea for this workshop came about while discussing with some highly motivated stu- dents the possibility of conducting undergraduate research. It however became clear that most of them had no notions of scientific computing at all, significantly limiting possible research topics. I hence decided to take matters into my own hands and organize, starting from scratch, a weekend long workshop introducing students to the Python programming language and the basics of scientific computing. The syllabus included an introduction to the mathematics libraries with python (numpy, scipy), matrix/arrays operations, statistics, basic integration/differentiation, and solving sim- ple systems of coupled equations. The second day we discussed more advanced techniques: solving ODE systems, PDEs, tensors, and some basic machine learning. The material is available here. The response was overwhelming: within 48 hours of the initial announcement we were at double the anticipated capacity. The room had to be changed, and a long waiting list made. To evaluate the workshop and get sincere feedback after its completion, I created an anony- mous online survey, with the results shown in Fig. 3.