101 points by quantumqueen 1 year ago flag hide 18 comments
quantum_researcher 1 year ago next
Exciting to see progress in quantum machine learning! I've been working on implementing KMeans on a quantum computer.
quantum_researcher 1 year ago next
It varies depending on the dataset size and dimensionality, but we see improvements around 10x for large, complex datasets.
quantum_chemist 1 year ago next
Amazing to see quantum computing provide speedups for algorithms with real-world applications. I've been working with quantum computers to tackle computational chemistry problems.
ai_enthusiast 1 year ago prev next
How much of a speedup are you seeing for KMeans compared to classical computers, quantum_researcher?
theoretical_quantum 1 year ago next
This dataset from Google might provide a compelling case study: https://quantumai.google/quantum-learning/datasets/taxi
another_user 1 year ago prev next
Fascinating! What quantum hardware are you using for the machine learning algorithms, quantum_researcher?
quantum_researcher 1 year ago next
We've been fortunate to have access to IBM's Q System One for these experiments. Excited for hardware like this to become more widely available.
in_the_know_quantum 1 year ago prev next
Just a reminder that there's always the risk of decoherence impacting performance on quantum computers, which can negate some benefits.
deep_learning_dev 1 year ago prev next
I've been comparing vanilla machine learning algorithms to their quantum counterparts. It's remarkable how similarly the two classes of algorithms learn from the data.
quantum_researcher 1 year ago next
Fascinating! We should collaborate and submit a paper showing how quantum machine learning can be used as a drop-in replacement for classical algorithms.
ai_enthusiast 1 year ago prev next
So, you're indicating that quantum machine learning algorithms can be a drop-in replacement for classical algorithms, deep_learning_dev? How generalizable is this statement?
deep_learning_dev 1 year ago next
Pretty generalizable if the quantum machine learning algorithms can be scaled with dataset size. This will be critical, as it will allow us to handle the increasing amounts of data we collect.
classical_guy 1 year ago prev next
While the work is exciting, mainstream business applications are likely still years away. Classical computers can handle the majority of needed tasks today.
another_user 1 year ago next
Yes, but wouldn't it be advantageous to start and familiarize oneself with the technology early? Even Toyota is working with quantum computers in their research now.
quantum_researcher 1 year ago prev next
classical_guy, I agree with your assessment. However, like another_user said, the early-mover advantage is real. We don't need quantum machines in every business right now, but we need researchers to keep pushing the technology forward.
quantum_optimist 1 year ago prev next
Quantum computers will revolutionize how we solve problems, from machine learning to cryptography and beyond. Decades of work will prove this out, and the world will change for the better.
quantum_skeptic 1 year ago next
It's an interesting field, but I'll be holding back on saying it will revolutionize our world. Quantum supremacy is still a bit further away than people think.
quantum_optimist 1 year ago next
While I understand your concern, it's important to remember that even small steps taken in the right direction can result in fantastic progress. Quantum supremacy might be further away, but we are getting there.