Journal

It was a great day to celebrate along Dominik and Raj!

After a short presentation and set of thought provoquing questions from the examiners I was granted a PhD subject to minor ammendments.

I'm thankful for the time the examiners invested in going through my thesis, as this resulted in very interesting questions that allowed me to enjoy this moment of conclusion.

I have been selected by my Research Centre as their candidate for Best PGR of 2023. A great honour!

I was invited to give a talk in the PhD students seminar and I used this opportunity to share with them the main ideas behind CAD, the algorithms I've been working on, and its multiple applications.

I left the city that has been my home for more than three years, leaving behind multiple friends, learnings and great experiences.

ISMA Coventry Festival 2023 is a festival at Coventry University where we were able to experience Indonesian music, dance and taste, specially that of Es Teler.

Alicia Dickenstein, a world-renowned expert in Algebraic Geometry and its applications to biology, came to give a talk at the Warwick University. Thanks to the proximity I was able to attend his talk on “Iterated and Mixed Discriminants” together with my second supervisor.  

During McCallum's visit to Coventry I had the opportunity to learn more about him and his supervisor Collins, the creator of CAD.
I had the opportunity to carefully present him my latest attempts to prove that clustering can be safely utilised while building a CAD. He gave me good feedback and pointed me to some good references that hopefully soon will result in a finished proof.

On this day, I had the opportunity to finally meet Scott McCallum, author of the latest improvements in CAD theory.

I presented my latest ideas regarding how clustering in CAD might be an absolute game changer and a bit more adventurous idea on how to create an order-invariant CAD under all circumstances, including nullification, something never done before.

Thanks to funding from the EU COST action, EuroProofNet, we received a visit from the PGR researchers  Jasper Nalbach (RWTH Aachen, Germany) and Lucas Michel (U. Liege, Belgium) during which we discussed about CAD theory and how existing implementations could be improved.

During the SC^2 workshop I presented my work on "Augmenting mathematical datasets" that resulted in a great improvement on the performance of machine learning models to pick a variable ordering for Cylindrical Algebraic Decomposition.

I have been looking forward to sharing this side project for over a year now. 

I created a poster that helped me explain to everyone in ISSAC the main idea of the improvement I propose.

It was great to see people following my explanations and I received the best-poster award.
🥳 

More details can be checked here. 

Great day in Bath discussing Cylindrical Algebraic Decomposition, my fellow PhD student in Bath, Corin, is facing the same problems I was facing a year ago. Hopefully, my code and comments will be useful for him, and we may end up creating some code to draw CADs.

I'm very happy to live next to one of the best universities in the UK in mathematics. Moreover, I just had the opportunity to talk to them about how Cylindrical Algebraic Decomposition can be useful to them and about the progress I've made speeding up this algorithm.
Hopefully, I can attend more seminars and colloquiums at Warwick in the future.

I recently developed a software that recommends darts players where to aim depending on their skill and the position of the game.
Pitching for the Dragons allowed me to make the software more user-friendly.

Lucian Moraru worked on "Training an agent to play a board game using reinforcement learning". 

He developed an agent to play The Game that was able to beat me the only time I played against it.

During this week we have been able to learn multiple useful things for PhD students, I just wish we had this event earlier in my PhD journey.
Some aspects need improvement for future editions, but I definitely know a lot more about publishing, originality, rigour, significance and writing. And in my opinion more importantly I got a couple of new ideas on how to improve the way I organise myself.

During this semester I have been supervising Arjun Jayan with his master's project. He has been working on "Bank Loan Eligibility prediction with explainability using machine learning approaches".
He build an app, that would not only tell you whether you are eligible for a loan or not, but also, informs you about what was the main reason for your rejection according to the explainability done on the machine learning model. Not only that, but it also suggests a way of changing your data (credit score, salary...) in a way that would guarantee acceptance of the loan.

It was very enjoyable to participate in this event, and working on solving logic&math problems under time pressure. There participation was not as high as I was expecting but that helped me secure the first position in the General Public category 😃
I also had the opportunity to meet Mike Paterson (Emeritus Professor at University of Warwick) and Ben Pridmore (former world memory champion) among other amazing people.

During the open problem session, I lead the group working in proving that the number of parking functions of length n adding up to k are uni-modal.

I also had the opportunity to attend Valborg in Uppsala, a very enjoyable festival in a very beautiful day!

Together with Rashid Barket, I attended this event to learn about different methods of data generation in mathematics and to present our own. Not only we did do that but we also learnt a ton about automatic theorem proving, that is the idea that took me into coding in the first place.
We also learnt about the COST project and we are hoping to strengthen our links with them in the future.

Together with Dominik, I organised the first internal conference in our centre.
The conference was a success and we were able to learn about each other's projects and the methodologies each of us are following.

After a couple of beers, I accepted the challenge of running a marathon in under five hours. It's a reasonable time that allows you to walk most of the time.

I made some strategic mistakes at the beginning and in the last kilometres my strength disappeared, but with the indispensable assistance of Rashid, I managed to complete the quest.

In this workshop, I teamed up with Hamid and Corin to build the first implementation of open Cylindrical Algebraic Decomposition in Macaulay2.

It has been a challenge and a great experience that has taught me a lot about how to develop code together with a team.

I want to thank the organisers for putting together this event in the Mathematics department of Warwick uni, a brilliant place where everyone is discussing maths around you. 

Thanks to the Research Excellence grant, I have been able to visit Berk Yıldız to work on trying to prove the log-concavity of a sequence in parking functions.
It was an honor to work in the Berlin Mathematical School, one of the most prestigious mathematical institutes in Europe were Maryna Viazovska (Field medallist) was working some year ago.
Even though we didn't achieve our dream goal we have been able to the number of descents in teleporting parking functions and we propose a faster was of computing the relevant sequence.

Our beloved Operations Support Manager Joel Gibbs bought us a dartboard and we have been having amazing breaks since.
With the support of the CSI-COP project, we organised a little tournament between staff and students. My team, "The Stingers", ended up winning and we got a nice 3D-printed trophy.
Thank you guys for the amazing event!

While Lynn stay in Coventry we started collaborating on a paper. Thanks to the Research Excellency grant I have been able to visit Lynn during in Gent, where she's working right now. We made much more progress than in the last months working together online, and now the paper is almost ready for submission.
Not only that but they also got to enjoy the beautiful city of Gent, its bikes and its cafes!

During this event, my group worked in collaboration with the AMRC on a dataset of expensive and rare parts of planes. The challenge was to augment an industrial dataset of expensive aeronautics components to train a model to spot defects as early as possible during manufacturing.

We encountered many difficulties but it was a pleasure working with my group and having fun during this amazing week organised by the Turing Institute.

At CASC I had the opportunity to meet many researchers in my area and to learn about a couple of very interesting problems related to graph domination and teleporting parking functions. I gave a talk on "New Heuristic to Choose a Cylindrical Algebraic Decomposition Variable Ordering Motivated by Complexity Analysis"

This seminar gave me the opportunity to discover an amazing place called Dagstuhl and to put faces to the authors of the papers I have been focusing on during my PhD. I was one of the VSA's and I gave a talk showing many possible uses of Machine Learning in the context of Satisfiability Checking and Symbolic Computation.

A great trip to finally met everyone in person.