Hi Wei, Thanks for your introduction, what I prefer to see is that if there's a proposal of what you are going to do in the 12 weeks of the period and it will be best to have written a doc explaining the different solution you have considered and why you end up with how it will be implemented. And it will be nice if there's a draft PR in your local fork showing us how you are going to implement it. At last I hope you can make a appearance in the Mahout Community Sync, it will help us to get to know you better.
Best regards, Jie-Kai On Wed, Mar 25, 2026 at 4:47 PM Leong Wei Chan <[email protected]> wrote: > > Hi Ryan and Mahout team, > > I hope you’re doing well. My name is Wei, and I’m very interested in > contributing to the “Add ZZFeatureMap Encoding for QDP” project for GSoC 2026. > > I’ve been working on quantum machine learning (at Scenario X), particularly > on kernel-based methods and feature mappings (on financial data), and the > ZZFeatureMap is something I’ve encountered frequently in Qiskit. I’m > especially interested in how different encodings affect expressivity and > model performance, so this project immediately stood out to me. > > I want to be upfront that while I have experience with quantum ML concepts > (e.g. feature maps, kernels, simulation workflows), I have not yet > contributed to a production-level system like QDP. That said, I’m very > willing to learn quickly and put in the work to understand the codebase and > design decisions in Mahout. > > To get started with this topic, I’ve been exploring the QDP encoding > implementations in the Mahout repo (particularly the amplitude, angle, basis, > and IQP encodings under the QDP encoding modules), and reviewing how > ZZFeatureMap is implemented in Qiskit’s circuit library (documentation). > > Thank you for your time, and I’d really appreciate any guidance and > mentorship you can offer. I’m excited about the possibility of contributing > to Mahout and learning from the team. > > Best regards, > Wei
