The Qiskit Challenge India Proved the Best Way to Learn is By Doing
By Ryan F. Mandelbaum, Senior Technical Writer, IBM Quantum and Qiskit
From the moment we announced the Qiskit Challenge India, we said that the best way to learn is by doing. This held true, not just for the hundreds of people who participated in and completed the challenge, but even for those on the Qiskit team who worked to create and support it.
Over 1100 quantum novices registered for the two-week challenge, which began on September 1. The first week consisted of readings followed by a series of questions about the basics of quantum computing, and those who completed the first week received a certificate. But for the second week, participants tackled a problem in the cutting-edge field of quantum machine learning. The Qiskit Challenge India demonstrated that in a new field like quantum computing — quantum machine learning, especially — sometimes the only way to get started is simply to plunge into the deep end.
“I want people to know that there is a lot of interest in the field, and that now is the right time to get into it — even if you have minimal knowledge of quantum computing,” said Rana Prathap Simh Mukthavaram, Co-op at IBM Quantum and Qiskit who devised the challenge.
The challenge kicked off with a quick refresher on quantum computing’s fundamental machinery: the mathematical tools of linear algebra. The pace picked up quickly, forcing participants, 67% of whom had never interacted with Qiskit before, to figure out which gate would produce a given quantum state, entangle qubits, and use gates to modify those entangled qubits.
But perhaps the real challenge began on Tuesday, September 8th. Awaiting participants was a behemoth of a Jupyter Notebook detailing the nitty-gritty of classifying data with a quantum computer, as well as a challenge question: write a quantum program that could tell the difference between the number 4 and the number 9 in a dataset of written-out numbers. Participants would have to combine all of the Qiskit knoweldge they gleaned from the previous week of problems, while creating the best possible solution to this open quantum machine learning exercise.
Indeed, all of the “doing” really did lead to “learning.” Pragya Katyayan, doctoral student in the Department of Computer Science at Banasthali Vidyapith in Rajasthan, India, started the challenge with a machine learning background, but was a novice at quantum machine learning, including variational circuits, feature maps, and quantum computing in general. By the end of the challenge, she not only understood these concepts, but could use them on new datasets. Her team’s submission ended up in the top 100.
“I strongly feel that this challenge was all I was looking for,” she said. “I had been trying to hard to understand the variational quantum classifier (VQC) and how QML works, but that struggle met its end withthis challenge. I am happiest that I became a part of the challenge and learned the concepts of VQC by actually playing with it.”
All the while, the Qiskit team was learning on their own — about the Qiskit community in India, about how to run such a large competition, and how to keep people engaged. The mentors would notice that certain concepts were especially difficult for the challenge participants, and huddle in order to create new content and FAQs to distribute. In order to keep participants active over the weekend, the team dropped a bonus question. The team applied startup-style thinking to act quickly and iterate in order solve any snags that arose.
Participants learned things beyond quantum concepts as well. Sure, BITS-Pilani undergraduate Prajjwal Vijaywargiya learned more during the challenge than he would have just sticking to the Qiskit textbook on his own. But he also took the initiative to meet new people in order to form a team — a personal victory, he said. “My biggest successes were the connections I made with people in this community, which, everyday since the challenge, has motivated me to keep my learning going and to someday actively contribute to the development of quantum computing.”
Rahul Pratap Singh, Co-op at IBM Quantum and Qiskit, hopes that more people will learn quantum computing by trying it out — after all, the field is so new that anyone can get involved and provide new perspective on what benefits the technology might have, some day. Meanwhile, the Qiskit team will continue encouraging folks to join the quantum computing community so we can continue to learn by doing through similar engagements and events.
After all, said Rana, “This field is going to blow up at some point. It’s just a matter of time, and learning by doing is a way to getting ready for it.”
Interested in joining the Qiskit community? Check out Qiskit.org to get started.
