Challenges
B
-
Barclays
- Challenge number 23
Creation of Synthetic Financial Data
Develop mathematical algorithms to create realistic synthetic financial datasets that preserve the statistical properties and complex correlations while ensuring privacy guarantees. The algorithm needs to represent primarily transactional data and should scale to hundreds of columns
- Challenge number 23
-
bp
- Challenge number 26
22. Quantum Optimisation for a Holistic, Multi-Hub UK Logistics Network: Routing, Disruption Replanning, Utilisation and Carbon Reduction, 24. 10. How should optimisation problems be formulated when uncertainty is intrinsic to the objective, constraints, or outcomes rather than treated as a post hoc analysis?, 9. What steps should industry be taking today in order to prepare for any emerging standards? What should industry credible evaluation standards cover to effectively evaluate emerging computational methods compared to established techniques?
22. [see full challenge attached]
Can quantum and hybrid classical-quantum methods solve the multi-objective optimisation problem of a holistic, multi-hub UK logistics network - simultaneously routing vehicles, replanning in real time after disruptions, maximising container utilisation, minimising unproductive driver time, enabling drop-and-swap hub operations and reducing carbon emissions - where the coupling of these objectives across hundreds of nodes makes classical decomposition progressively inadequate?
24. 10. Optimisation Under Uncertainty: Risk Aware Problem Formulations
In many decision problems, uncertainty is intrinsic to objectives, constraints, or outcomes rather than something that can be analysed after optimisation. In other words, in a real-world setting, uncertainty in the objectives, constraints, or outcomes will always be present, and so an optimisation will only be useful if the optimum is robust to fluctuations in these inputs. Treating uncertainty as a first class component of the formulation often leads to different solution strategies and computational requirements. However, there is limited consensus on when and how to use robust, stochastic, or risk aware formulations in practice.
9. Standardising the Evaluation of Emerging Computational Methods
As new computational methods emerge, organisations struggle to evaluate them consistently against established techniques. Claims of performance improvement are often difficult to compare, reproduce, or interpret across studies. There is a need for shared evaluation standards that are method, solver, and platform agnostic while remaining practically applicable. Standards networks and initiatives are emerging, but to gain full benefit from these, industry may still need to do some independent preparation – for example by adopting relevant evaluation approaches in existing classical methods that allow comparison.
- Challenge number 26
D
-
Digital Catapult
- Challenge number 17
How do we best connect researchers and education materials with industry?
We engage with end users and often try to connect them with academia, this is done on a very case by case basis. As industry becomes more and more engaged, can we streamline this across the quantum ecosystem to truly prepare the future quantum workforce.
- Challenge number 17
E
-
ENODA Ltd.
- Challenge number 7
Quantum optimisation for distributed flexibility dispatch: As electricity systems become more decentralised, energy aggregators increasingly manage portfolios of distributed energy resources across many sites, including batteries, flexible industrial loads, and commercial consumers. These resources can provide multiple balancing services, such as aFRR up and aFRR down, but their operation is constrained by power limits, energy capacity, availability, recovery requirements, contractual commitments, and site-level restrictions. Decisions made in one settlement period also affect what is possible and economically optimal in later periods. How can quantum or quantum-inspired optimisation methods support the scheduling of distributed energy resources across multiple sites, services, and time periods, while respecting real-world operational constraints?
Background and context:
Aggregators must decide which assets should participate in which flexibility products, in which direction, at which site, and at what time. For storage assets, charging has an energy cost and changes the state of charge available for future dispatch. For demand-side assets, increasing or decreasing consumption may have operational costs or recovery effects. As the number of sites, resources, products, and time intervals grows, the number of possible dispatch combinations increases rapidly.
Classical optimisation methods can solve many practical scheduling problems today, but large multi-site, multi-product portfolios may become computationally challenging when uncertainty, recovery constraints, market prices, and asset interactions are included. This raises the question of whether quantum approaches, quantum-inspired heuristics, or hybrid classical–quantum methods could offer useful speed, scale, or solution-quality advantages for future flexibility markets.
- Challenge number 8
Quantum methods for probabilistic voltage headroom in distribution networks: Distribution networks are increasingly affected by uncertain demand, distributed generation, electrification, and active voltage control. Network operators and flexibility providers need to understand how much voltage modulation is possible before statutory voltage limits are breached, and how much demand reduction can be achieved with high confidence. This requires reasoning over network topology, power-flow constraints, uncertain consumer behaviour, and the non-linear relationship between voltage and demand. Can quantum computing methods improve the calculation of probabilistic voltage headroom in distribution networks, and is there a route to genuine quantum advantage for this type of constrained uncertainty problem?
Background and context:
Consider a distribution grid with a known topology and admittance matrix, connected to many consumers whose future demand or generation is uncertain. Historical smart meter data or machine learning models may provide probability distributions for consumer behaviour, but the operator still needs to determine whether a proposed voltage modulation can reduce demand while keeping all network voltages within statutory limits.
A practical method would need to estimate a feasible range of voltage modulation values that achieves a required demand reduction with a specified confidence level, for example 90%, while ensuring voltage constraints are not violated across the network. Classical approaches may involve repeated power-flow calculations, Monte Carlo simulation, or probabilistic optimisation. This creates an opportunity to explore whether quantum methods, such as quantum-accelerated Monte Carlo, quantum optimisation, or hybrid approaches, could provide a meaningful advantage over classical techniques.
- Challenge number 11
Quantum optimisation for energy device geometry design: The performance of advanced energy devices can depend strongly on their geometry. In the case of the ENODA Prime Exchanger, the shape and distribution of ferromagnetic limbs, together with the geometry and placement of coils, can significantly affect the device’s energy manipulation capabilities. Finding high-performing configurations requires searching across a complex design space with many interacting parameters. Can quantum or quantum-inspired optimisation methods improve the search for high-performing energy device geometries, particularly where the design space is non-convex, non-smooth, or difficult for traditional optimisation methods?
Engineering design problems often involve complex optimisation landscapes. While many classical optimisation methods work well for smooth, convex, differentiable functions, real-world device design can involve non-convex objectives, discontinuities, simulation-based evaluations, manufacturability constraints, and strong interactions between design parameters. These characteristics can make exact optimisation impractical and heuristic search slow or incomplete.
This challenge asks whether quantum optimisation, quantum-inspired heuristics, or hybrid classical–quantum approaches could help explore the geometry design space more efficiently. The focus is on whether these methods could improve the optimisation of exchanger geometry and coil placement, either by accelerating the search process, improving solution quality, or expanding the types of engineering design problems that can be practically optimised.
- Challenge number 7
F
-
Financial Conduct Authority
- Challenge number 18
Coordinating the ecosystem to overcome barriers to quantum readiness in financial services
The transition of quantum computing from labs to real-world applications in financial services present challenges that extend beyond technological breakthroughs. While advances in hardware, algorithms, and tooling remain critical, many of the barriers to adoption are fundamentally ecosystem challenges involving collaboration, trust, skills development, governance, and shared understanding between stakeholders.
This challenge explores how financial institutions, academia, quantum vendors, and the public sector can work together to overcome the organisational, co-ordinational, cultural, and operational barriers that currently limit quantum readiness and adoption.
This AIMday workshop will utilise User-Centred Design methods to bring together stakeholders from across the ecosystem including end users, academia, quantum start-ups, technology providers, and regulators to explore:
The barriers they currently face;
Their goals, incentives, and requirements; and
What more coordinated action may be required to accelerate readiness, adoption, and deployment.
The session aims to identify practical insights and opportunities for ecosystem collaboration that can help support the meaningful adoption of quantum computing in financial services.
- Challenge number 18
H
-
Health Data Research UK
- Challenge number 28
2. Investigating using Quantum Computing and Hybrid Quantum Machine Learning for bioinformatics., 27. 13. Are quantum machine learning methods suitable to cluster healthcare data?, 1. Harnessing data for quantum-engaged systems in therapeutics for healthcare
2. QBA is a SME, based in Brighton and the Institute of Physics Accelerator London ,focussing on industrial uses for quantum computing (QC) and AI.Our project Q-Bio MERS explores the benefits of using Quantum Machine Learning (QML) and advanced classical AI for improving medicine discovery specifically by studying the genome of candidate pathogens starting with the virus MERS (Middle Eastern Respiratory Syndrome.)
The UK Vaccine Network advising the DHSC have identified priority pathogens in 12 viral and bacterial families including the Coronavirus family (e.g. MERS) and Yersinia pestis (the Plague)
Furthermore, it seeks to establish a general technological approach to targeting future Disease Xs (World Health Organisation placeholder title for future pathogenic epidemics.)
27. 13. Healthcare data tend to be high dimensional. In order to load the data into a quantum computer, it is necessary to perform "aggressive" dimensionality reduction that would not be necessary when using classical methods, in doing so information and subtle but important correlations might be lost. In addition, a popular method to perform clustering seems to be quantum kernel methods but they are not scaling well, especially if using quantum kernel alignment (QKA) that also comes with potential Barren Plateau. In light of that, can performing QKA on a small randomly selected subset and then generalize the clusters to the other datapoints a viable approach?
1. Applications of quantum computation are breaking through into therapeutics in healthcare. For example, Algorithmiq recently won a £2m dollar Wellcome Leap prize for their work applying end-to-end quantum-classical algorithms to simulate the activation pathway of a photosensitiser drug, currently in Phase II clinical trials (see https://algorithmiq.fi/news/algorithmiq-wins-2-million-wellcome-leap-prize-for-quantum-enabled-cancer-drug-discovery-development/). In systems like this, the quantum computation is an element of a more complex pathway (in drug develpment or in other complex healthcare areas) that integrate genomic/phenomic data with (perhaps) complex simulations at molecular level. This raises the issue of where best to situate the quantum computation, for example in the molecular chemistry (as in the Algorithmiq application) or in representing genomic information (e.g. in the Sanger Institute’s “loading’ of the Hepatitis D viral gene into a quantum computer https://www.sanger.ac.uk/news_item/genome-loaded-onto-a-quantum-computer-in-world-first/). How do we create the data and process management systems that enable many future quantum-linked drug discovery systems to be rapidly established, configured and securely applied?
- Challenge number 28
I
-
Infleqtion
- Challenge number 12
Is it worth implementing codes that offer high encoding rate thanks to non local connectivity?
As we can move atoms, we can implement codes that provide high encoding rate thanks to non local connectivity such as the bivariate bicycle code. This is ideal in theory but in practice, moving atoms takes time and increases atom loss. How can we balance the cost of moving atoms with the high encoding rate? Is there a combination of codes that can be used to get the "best of both worlds"?
- Challenge number 14
What prospects remain for demonstrating variational quantum advantage on near-term hardware?
Hybrid quantum-classical algorithms based on variational principles, whether for preparing physically-relevant states for quantum simulations (VQE), or solving combinatorial optimisation problems (QAOA), have dominated the last decade of NISQ applications research. The initial excitement stemming from successful implementations on shallow problems across numerous disciplines, while deserved, has waned in recent years under the growing realisation that "barren plateaus", the exponential-vanishing of cost-function gradients, are quite generic features as one's problem dimensionality grows.
At the same time, a panoply of richer variational ansatz have become available over the last five years, including those constrained by the symmetries of the optimisation target, those constructed adaptively to the problem (e.g. ADAPT-VQE), and those building on machine learning architectures (e.g. NN-VQE). Many of these designs are guided, in part, by the novel analytical understandings of barren plateaus and their avoidance, and have made possible 100+ qubit simulations of relevant lattice quantum systems. Similarly, warm-start procedures which restrict to the "dynamical Lie algebra" of a VQE ansatz have shown promise in avoiding barren plateaus.
Taking a broad view across quantum simulation and combinatorial optimisation, on what strategies should frontier hardware labs centre their focus if they want to produce credible demonstrations of variational quantum advantage? Indeed, should labs deprioritise variational algorithms in favour of more explicitly scalable demonstrations, or in doing so, are they missing clean near-term opportunities to showcase their hardware capabilities using state-of-the-art ansatz design, warm starts, and error-mitigation strategies? What evidence would make such a variational demonstration convincing to end users, and which algorithms or ansatz are best suited to providing that evidence?
- Challenge number 28
2. Investigating using Quantum Computing and Hybrid Quantum Machine Learning for bioinformatics., 27. 13. Are quantum machine learning methods suitable to cluster healthcare data?, 1. Harnessing data for quantum-engaged systems in therapeutics for healthcare
2. QBA is a SME, based in Brighton and the Institute of Physics Accelerator London ,focussing on industrial uses for quantum computing (QC) and AI.Our project Q-Bio MERS explores the benefits of using Quantum Machine Learning (QML) and advanced classical AI for improving medicine discovery specifically by studying the genome of candidate pathogens starting with the virus MERS (Middle Eastern Respiratory Syndrome.)
The UK Vaccine Network advising the DHSC have identified priority pathogens in 12 viral and bacterial families including the Coronavirus family (e.g. MERS) and Yersinia pestis (the Plague)
Furthermore, it seeks to establish a general technological approach to targeting future Disease Xs (World Health Organisation placeholder title for future pathogenic epidemics.)
27. 13. Healthcare data tend to be high dimensional. In order to load the data into a quantum computer, it is necessary to perform "aggressive" dimensionality reduction that would not be necessary when using classical methods, in doing so information and subtle but important correlations might be lost. In addition, a popular method to perform clustering seems to be quantum kernel methods but they are not scaling well, especially if using quantum kernel alignment (QKA) that also comes with potential Barren Plateau. In light of that, can performing QKA on a small randomly selected subset and then generalize the clusters to the other datapoints a viable approach?
1. Applications of quantum computation are breaking through into therapeutics in healthcare. For example, Algorithmiq recently won a £2m dollar Wellcome Leap prize for their work applying end-to-end quantum-classical algorithms to simulate the activation pathway of a photosensitiser drug, currently in Phase II clinical trials (see https://algorithmiq.fi/news/algorithmiq-wins-2-million-wellcome-leap-prize-for-quantum-enabled-cancer-drug-discovery-development/). In systems like this, the quantum computation is an element of a more complex pathway (in drug develpment or in other complex healthcare areas) that integrate genomic/phenomic data with (perhaps) complex simulations at molecular level. This raises the issue of where best to situate the quantum computation, for example in the molecular chemistry (as in the Algorithmiq application) or in representing genomic information (e.g. in the Sanger Institute’s “loading’ of the Hepatitis D viral gene into a quantum computer https://www.sanger.ac.uk/news_item/genome-loaded-onto-a-quantum-computer-in-world-first/). How do we create the data and process management systems that enable many future quantum-linked drug discovery systems to be rapidly established, configured and securely applied?
- Challenge number 12
L
-
Lloyds Banking Group
- Challenge number 19
Stress-testing quantum safe algorithms.
As a bank we have requirement to move payments quickly. Moving to a quantum-safe algorithm to protect payments could put this requirement at risk. Can you provide stress-testing on the hardware performance and response latency of payments with quantum-safe algorithms to demonstrate how we can meet this requirement while still providing the most secure service?
- Challenge number 19
M
-
M&G Investments
- Challenge number 21
Scaling Portfolio Optimisation to High-Dimensional Parameter Spaces
[please see attachment for fuller description and objectives/aims]
M&G’s portfolio optimisation capability is under increasing strain as investment strategies incorporate a broader set of data, risk measures, and constraints. This expansion is driving a significant increase in model complexity, creating a gap between the volume of available information and the ability of classical optimisation techniques to fully utilise it.
As additional parameters are introduced - such as ESG factors, alternative data, and non-linear constraints - the optimisation problem becomes high-dimensional and computationally intractable. Current approaches rely on simplifications and dimensionality reduction to remain tractable, but these inherently limit the solution space and constrain the ability to maximise risk-adjusted returns.
This proposal evaluates whether quantum and hybrid classical–quantum approaches can address these limitations by enabling scalable optimisation across significantly larger parameter sets. The expected outcome is a clear, evidence-based view on the viability of quantum techniques, how problems should be structured for them, and where they may deliver practical value - informing future investment decisions and positioning M&G to exploit emerging computational capabilities.
- Challenge number 21
N
-
Nu Quantum
- Challenge number 25
3. Is space-time volume the best metric for a quantum resource estimate?, 4. Which applications will most significantly benefit from scaling up the number of qubits?
3. Resource estimates for quantum algorithms are often quantified in terms of the space-time volume: the number of physical qubits multiplied by the number of quantum error correction rounds. This is on the basis that these two quantities are inversely related: increasing the number of qubits will lead to a decrease in the time required for a quantum computation. What other resources could we consider when estimating resources, and how do these relate to the space-time volume? In particular, how might the power consumption of a fault-tolerant quantum computation vary with space and time?
4. Quantum circuits have a linear trade-off in terms of time and space: an n-qubit quantum circuit can be implemented with 2*n qubits in approximately half the time. This relation becomes more complicated when error correction overheads are considered: recent resource estimates for Shor's algorithm for instance have found that increasing the qubit count can lead to a non-linear reduction in the runtime (see arXiv:2602.11457). Are there other problems which also show this non-linear relation, and if so how can we optimise for both time and space when trying to solve these problems on a quantum computer?
- Challenge number 25
O
-
One World
- Challenge number 26
22. Quantum Optimisation for a Holistic, Multi-Hub UK Logistics Network: Routing, Disruption Replanning, Utilisation and Carbon Reduction, 24. 10. How should optimisation problems be formulated when uncertainty is intrinsic to the objective, constraints, or outcomes rather than treated as a post hoc analysis?, 9. What steps should industry be taking today in order to prepare for any emerging standards? What should industry credible evaluation standards cover to effectively evaluate emerging computational methods compared to established techniques?
22. [see full challenge attached]
Can quantum and hybrid classical-quantum methods solve the multi-objective optimisation problem of a holistic, multi-hub UK logistics network - simultaneously routing vehicles, replanning in real time after disruptions, maximising container utilisation, minimising unproductive driver time, enabling drop-and-swap hub operations and reducing carbon emissions - where the coupling of these objectives across hundreds of nodes makes classical decomposition progressively inadequate?
24. 10. Optimisation Under Uncertainty: Risk Aware Problem Formulations
In many decision problems, uncertainty is intrinsic to objectives, constraints, or outcomes rather than something that can be analysed after optimisation. In other words, in a real-world setting, uncertainty in the objectives, constraints, or outcomes will always be present, and so an optimisation will only be useful if the optimum is robust to fluctuations in these inputs. Treating uncertainty as a first class component of the formulation often leads to different solution strategies and computational requirements. However, there is limited consensus on when and how to use robust, stochastic, or risk aware formulations in practice.
9. Standardising the Evaluation of Emerging Computational Methods
As new computational methods emerge, organisations struggle to evaluate them consistently against established techniques. Claims of performance improvement are often difficult to compare, reproduce, or interpret across studies. There is a need for shared evaluation standards that are method, solver, and platform agnostic while remaining practically applicable. Standards networks and initiatives are emerging, but to gain full benefit from these, industry may still need to do some independent preparation – for example by adopting relevant evaluation approaches in existing classical methods that allow comparison.
- Challenge number 26
P
-
Protostars
- Challenge number 5
Quantum-inspired Watermarking for AI model software CE Marking
Protostars is building a Cybersecure AI continuous assurance platform that converts technical evidence (code, cloud configuration, identity, logs, SBOM/AIBOM and supply-chain artefacts) into regulator-ready evidence packs to support CRA, NIS2 and AI governance. The core problem we are solving is trust in evidence at scale: manual audits are too slow, self-attestations are weak as proof, and evidence is fragmented across tools and suppliers—yet regulators, auditors, insurers and certification bodies are all converging on the same question: “Can we trust the evidence?” This is becoming urgent as Europe shifts from point-in-time assessment to continuous evidence, with CRA and AI Act reporting obligations landing in the same period and “product is now the perimeter.”
A key technical gap is provenance and tamper-evidence for AI assurance artefacts. If an assurance report, SBOM/AIBOM, or model evaluation output can be copied, modified, or detached from the release that produced it, then governance becomes “theatre”. We therefore want to advance quantum-inspired watermarking as a “trust artefact” that binds AI outputs and evidence packs to signed build provenance (CI/CD attestations) and supports long lifecycle assurance. The research direction we want to take forward is inspired by work associated with the University of Edinburgh ecosystem (Quantum Software Lab), leveraging quantum-inspired optimisation to select robust embedding locations and to adapt embedding strength.
As a concrete baseline, the attached quantum-inspired watermarking method (in the literature) combines DWT + DCT and uses Quantum-Inspired Annealing (QIA) to optimise watermark embedding positions and Quantum Variational Circuits (QVC) to dynamically tune embedding strength (α) at block level. The reported results show high imperceptibility and robustness under common distortions (e.g., JPEG compression, noise, cropping), indicating the approach can produce watermarks that survive “real world handling.” Our challenge is to translate this from “image watermarking” into software and AI assurance watermarking: model artefacts (weights, configs, evaluation results), SBOM/AIBOM, audit reports, and evidence manifests used to support software CE marking and CRA cybersecurity-by-design/default evidence.
Finally, we are exploring whether quantum computing can help train sovereign AI models for the assurance domain—particularly models that can interpret technical evidence and generate structured compliance outputs for products and systems that include quantum components (e.g., quantum software stacks, hybrid classical/quantum workflows). In practical terms, we are interested in (i) quantum/quantum-inspired optimisation to improve watermark policy selection and verifier robustness, and (ii) how the quantum ecosystem can help define what “audit-ready, CE-relevant evidence” looks like for quantum computing products as Products with Digital Elements (e.g., provenance, secure update, vulnerability handling, and supply-chain integrity).
What we want from AIMday: expert input on (1) how to adapt quantum-inspired watermarking to AI model artefacts and software evidence packs, (2) how to design verifiers and attack models suitable for certification/supervision contexts, and (3) whether and where quantum resources (or quantum-inspired methods) can realistically accelerate sovereign assurance model training or assurance optimisation problems.
- Challenge number 5
Q
-
Quantum Base Alpha Ltd
- Challenge number 28
2. Investigating using Quantum Computing and Hybrid Quantum Machine Learning for bioinformatics., 27. 13. Are quantum machine learning methods suitable to cluster healthcare data?, 1. Harnessing data for quantum-engaged systems in therapeutics for healthcare
2. QBA is a SME, based in Brighton and the Institute of Physics Accelerator London ,focussing on industrial uses for quantum computing (QC) and AI.Our project Q-Bio MERS explores the benefits of using Quantum Machine Learning (QML) and advanced classical AI for improving medicine discovery specifically by studying the genome of candidate pathogens starting with the virus MERS (Middle Eastern Respiratory Syndrome.)
The UK Vaccine Network advising the DHSC have identified priority pathogens in 12 viral and bacterial families including the Coronavirus family (e.g. MERS) and Yersinia pestis (the Plague)
Furthermore, it seeks to establish a general technological approach to targeting future Disease Xs (World Health Organisation placeholder title for future pathogenic epidemics.)
27. 13. Healthcare data tend to be high dimensional. In order to load the data into a quantum computer, it is necessary to perform "aggressive" dimensionality reduction that would not be necessary when using classical methods, in doing so information and subtle but important correlations might be lost. In addition, a popular method to perform clustering seems to be quantum kernel methods but they are not scaling well, especially if using quantum kernel alignment (QKA) that also comes with potential Barren Plateau. In light of that, can performing QKA on a small randomly selected subset and then generalize the clusters to the other datapoints a viable approach?
1. Applications of quantum computation are breaking through into therapeutics in healthcare. For example, Algorithmiq recently won a £2m dollar Wellcome Leap prize for their work applying end-to-end quantum-classical algorithms to simulate the activation pathway of a photosensitiser drug, currently in Phase II clinical trials (see https://algorithmiq.fi/news/algorithmiq-wins-2-million-wellcome-leap-prize-for-quantum-enabled-cancer-drug-discovery-development/). In systems like this, the quantum computation is an element of a more complex pathway (in drug develpment or in other complex healthcare areas) that integrate genomic/phenomic data with (perhaps) complex simulations at molecular level. This raises the issue of where best to situate the quantum computation, for example in the molecular chemistry (as in the Algorithmiq application) or in representing genomic information (e.g. in the Sanger Institute’s “loading’ of the Hepatitis D viral gene into a quantum computer https://www.sanger.ac.uk/news_item/genome-loaded-onto-a-quantum-computer-in-world-first/). How do we create the data and process management systems that enable many future quantum-linked drug discovery systems to be rapidly established, configured and securely applied?
- Challenge number 28
T
-
Tata Consultancy Services
- Challenge number 20
Wind Speed Forecasting
Accurate short-term wind speed forecasting is essential for sectors like agriculture, energy,
transportation, and disaster management. Classical machine learning and statistical models (e.g.,
ARIMA, LSTM, Random Forest) have been widely used for this task. However, as climate patterns
become increasingly complex and datasets grow in size and dimensionality, classical approaches
face significant limitations. Current classical approaches face challenges like High
Computational Complexity, Non-linear Dependencies, Parameter Optimization Bottleneck,
Uncertainty Quantification, Scalability Issues
Recent studies indicate that quantum computing could effectively tackle problems of this nature,
offering benefits such as improved handling of high‑dimensional data, enhanced sampling for
probabilistic models, hybrid optimization efficiency, and future-ready scalability.
This is to explore if we could build a hybrid quantum-classical forecasting system that predicts the next 7 days wind speed for a given data.
Expected Output
The expected output is the day level wind speed for next 7 days.
Performance Metrics:
Choose appropriate ones
- Challenge number 20