Haiqu Quantum Gets Milestone In Quantum Machine Learning
Haiqu Demonstrates Quantum Machine Learning Efficiency on IBM Hardware, Signaling Near-Term Advantage in Anomaly Detection
Haiqu Quantum
A notable example produced by Haiqu Inc, a new quantum software startup, strongly argues that Quantum Machine Learning (QML) could soon bring practical benefits. The business established experimentally that current quantum computers are more successful than conventional classical systems at identifying patterns and detecting anomalies in huge, difficult datasets. IBM's powerful Quantum Heron enabled this anomaly detection breakthrough, an essential and resource-intensive global process.
The most convincing empirical indication to date that the promise of quantum advantage in data processing is quickly approaching the near-term is the successful implementation of quantum systems to handle the most complex aspect of data analysis, leading to increased accuracy and quicker preprocessing times over purely classical methods.
Bottleneck: Classical Limits and Dimensional Curse Modern infrastructure relies on anomaly detection, the “proverbial needle in the digital haystack”. It is needed to detect financial fraud, abnormal stock market transactions, tiny vital sign changes in patients, and unusual weather patterns.
In the Big Data era, data volume and complexity overwhelm standard algorithms. Real-world data is frequently categorized as “high-dimensional,” which means that hundreds or even thousands of attributes can be used to describe a single data piece. The “curse of dimensionality” refers to the exponential rise in computer resources required by classical systems to find important patterns or small outliers as the number of features rises.
Real-time analysis is crucial in high-frequency trading and real-time health monitoring, so this issue often causes operational bottlenecks. This can lead to costly false positives or, worse, missing detections. QML tries to take advantage of the fundamentally different representation and processing of information afforded by quantum computing in order to extract these subtle patterns more effectively than previous methodologies.
Haiqu: Quantum Embedding Scales QML Haiqu relies on a novel and effective quantum embedding approach. The bridge technology translates complex classical data into quantum computer-friendly format. A big classical dataset can be condensed into a sophisticated quantum circuit.
Its size distinguishes this proof-of-concept from others. Haiqu encoded more than 500 features from a complex financial dataset onto the IBM Quantum Heron processor's 128 qubits. This achievement marks a turning point because NISQ Noisy Intermediate-Scale Quantum, the previous practical limitation, could not load enough high-dimensional data to affect Quantum Machine Learning (QML) on existing quantum hardware.
Oleksandr Kyriienko, Professor and Chair in Quantum Technologies at Sheffield, stressed the technical significance. He stressed that quantum embedding defines model complexity and performance, therefore it must be understood and used while studying quantum device data. Professor Kyriienko said he was “very happy to see this implemented at an unprecedented scale,” adding that anomaly detection is a perfect target. Even a small score gain can lead to crucial detections or false positive eradication.
This excellent translation allows quantum applications to scale up, according to Haiqu's CTO and co-founder Mykola Maksymenko. Maksymenko believes quantum data processing can help with anomaly detection based on their research.
Faster Preprocessing and Better Accuracy
The hybrid quantum-classical technique was employed in the experiment. The most data-intensive step, preprocessing, was handled by the quantum computer. This quantum preprocessing step improved the high-dimensional financial data's feature set. An ordinary machine learning method was used to classify and detect anomalies from this quantum-enhanced feature set.
The quantum-enhanced preprocessing outperformed a pure classical baseline that employed random parameters to make a fair comparison. Quantum methods were more accurate at detecting irregularities in complex financial datasets.
The scientists also studied computing speed and discovered that preprocessing time on the genuine IBM Quantum Heron device was faster than when the equivalent operations were simulated traditionally. This significant observation suggests instant time savings for data preparation.
IBM Research Director Jay Gambetta complimented the study for encoding high-dimensional data with hundreds or thousands of characteristics, enabling new applications. Gambetta said “Advances like this are what push the industry towards achieving a quantum advantage in the near term”.
Signal, Not Claim: The Future
Haiqu's leadership is cautiously managing expectations of a quantum advantage despite impressive results. Haiqu's CEO and co-founder, Richard Givhan, said, “They are not claiming quantum advantage just yet.” However, he said they are providing the most convincing empirical evidence that (1) high-dimensional real-world data can be loaded onto a quantum computer and (2) QML may soon be useful for processing such data.
The latest research verifies past findings and improves reliability, control, and repeatability while offering more scalable embeddings and storing more classical data in quantum states. The approach was evaluated across many machine learning in ideal simulation and on hardware.
This technology could transform industries. Better fraud detection, risk modelling, and other applications reach beyond finance:
Healthcare: Monitoring minute medical readings may reveal health issues early. Industrial: Machine sensor failure detection for predictive maintenance.
Environmental monitoring: Faster detection of seismic anomalies like earthquakes.
Haiqu is accepting beta tester submissions to test their quantum feature embedding technique on these broader analysis challenges. The company expects the struggle for a quantum advantage to intensify when their quantum technique scales to address problems with tens of thousands of features on a near-term quantum processor.
