But that realization now leads me to increase the visibility of those offers and of my supervision activity in general, by also describing it here. In general, I consider myself a pretty good advisor - pardon my frankness and my full-of-s**tness. The reason I say so is that I have always put the interest of my students at the first place, much higher than the scientific goals and everything else. When a student asks me to work on a topic with me, I make sure I understand what short- and medium-term goals she has: is she planning to search for a Ph.D. after her masters thesis? Where does she sees herself in five-years time?
Depending on the answers I get, I am able to tailor the work plan of their thesis activities. So, e.g., for a person who wants to pursue a Ph.D. after the master, the authorship of an article is of paramount importance (it is one of the most valued bits of information on a CV in a Ph.D. selection); also of importance are presentations and posters at workshops and conferences. On the other hand, for a person who wants to find employment in the industry after graduation, I can offer some topics and research methodologies that strengthen her competences of value for the jobs market.
Over many years, I got a lot of very positive feedback from my former students (one of them even broke the rule "don't give gifts to your supervisor" by presenting me a fantastic bottle of Amarone, which I was really glad to accept and taste that very evening!). In one recent case, I picked up a Ph.D. student who had encountered difficulties with his former advisor, and led him to produce a thesis that was valued summa cum laude in less than 18 months - he thus commented on twitter after the fact:
Often, candidate students who are up for a master thesis advised remotely ask me if I have financial means to support their travel, or visits to my institute. The answer has always been the same: I have little research funds, and must use them wisely, but the biggest problem is that my institute makes it impossible for non-structured researchers to be funded as visitors. On the other hand, I tell them that there are possibilities connected with collaborations I entertain with other institutes (this did lead to the students visiting those places in a case or two), and with workshops, etcetera. And this is no joke - e.g., at the last MODE workshop in Crete one month ago we were able to invite a small cohort of students, all expenses paid, asking them in exchange to present a poster on their work. But most of all, it depends a lot on them! If they do valuable work, they make themselves attractive as future Ph.D. recipients, e.g.
So here I would like to use this space to advertise a few masters thesis topics that may be attractive to bright students in fundamental physics and/or ones who come from the field of computer science and want to pursue an applications-focused research. The list is continuously evolving, of course, but hopefully it will give readers the flavour of the things they can end up working on with me.
1) Muon Tomography and Deep Learning
In the context of the MODE collaboration we are developing a software package written in python (specifically with PyTorch) to create a fully differentiable model of a muon tomography imager. The purpose of muon tomography is to create 3D maps of unknown volumes using the detection of the trajectory of muons from cosmic rays that traverse the volume. The inference is an inverse problem (given the scattering angle, determine the material traversed), and can be tackled with CNNs or graph neural networks, but also with simpler expectation-maximization algorithms; but the point of producing a full model of the detector, the physics, the inference all together is to maximize a figure of merit - say, the precision of the imaging for given detector cost. The package, called TomOpt, is being developed by a team from institutes participating in MODE, and there are several activities within this project that lend themselves as perfect thesis topics. Publications are guaranteed, as well as authorship of a software which may one day be commercialized.
2) Muon Tomography by k-Capture
This is an idea of mine connected with the topic above, where one may use the physics of negative muon stops in dense materials to infer the detailed atomic properties of the absorbing material. I described this in detail in a recent blog post, so I will not repeat anything here, except to say that it is an innovative topic which could allow a master student (or even a PhD student, for that matter) to publish a source paper of relevance.
3) Assessing particle ID capabilities in a granular calorimeter
In large HEP experiments such as ATLAS and CMS at CERN, hadron calorimeters have become very important to detect substructure within the messy hadronic jets that particle collisions produce. The substructure may originate from decays of heavy particles (Higgs bosons, W,Z, top quarks) that are themselves potential flags of new physics signatures. The question then arises on how finely segmented it is advantageous to build our calorimeters. In this work we will simulate an impossibly thinly segmented calorimeter, subject it to fluxes of protons, pions, kaons, neutrons of fixed momenta, and proceed to discriminate their ID from the shape and detailed characteristics of their hadronic showers; then, once the "measurable discrimination" is assessed in this ideal instrument, we will study how and when does that information get lost as we increase the size of calorimeter cells. This will provide crucial information for the optimization of calorimeters for future colliders. Again, this work would ideally produce at least one or two publications in the time scale of relevance of a master thesis.
The three above are only a few of the topics available, and those less connected with explicit data analysis with the CMS experiment, of which I am a member. I had to make a choice here, and I decided that the "data analysis for HEP" topics are less special and interesting to broadcast. But of course, if you are interested there is a lot to do e.g. in studying the large datasets of B hadrons that CMS has so far produced with its "B-parked" datasets. One analysis I am involved in is the search for the neutral B_s meson in its decay to tau leptons, a process which might be sensitive to new physics contributions.
So, if you are interested, just drop me a message and we'll talk!