Alumni Projects - Graduation Year: 2024
Hydrodynamic Quantum Field Theory Analogs
Vishal Nandakumar, Pranav Panicker, Kai Wang
We explore the hydrodynamic quantum field theory, a model of quantum dynamics inspired by Louis de Broglie’s double-solution pilot wave theory and informed by the hydrodynamic pilot-wave system discovered by Couder and Fort in 2005. de Broglie originally proposed that every quantum particle contains an internal oscillation at the Compton frequency, exchanging its rest mass energy with its pilot wave field energy. de Broglie postulated that this pilot wave would satisfy the Klein-Gordon equation and Dagan and Bush have extended this theory by modeling the particle’s oscillations as localized disturbances to the scalar pilot wave field. We start by physically modeling the superwalking droplets and the pilot wave theory in a silicone oil bath. We extend the two-dimensional form of the hydrodynamic pilot wave to three dimensions by exploring the free particle, the harmonic oscillators, and other quantum analogs. We also explore the possible link between the non-Markovian dynamics of the physical pilot wave system and nonlocality in quantum systems.
Exploring Geometrical Properties of Chaotic Systems Through an Analysis of the Rulkov Neuron Maps
Nivika Gandhi, Brandon Le
Dynamical systems theory is a branch of mathematical physics with countless applications to numerous fields. Some dynamical systems exhibit chaotic behavior, characterized by a sensitive dependence on initial conditions commonly known as the "butterfly effect." While extensive research has been conducted on chaos emerging from a dynamical system's temporal dynamics, our research examines extreme sensitivity to initial conditions in discrete-time dynamical systems from a geometrical perspective. Specifically, we develop methods of detecting, classifying, and quantifying geometric structures that lead to chaotic behavior in maps, including certain bifurcations, fractal geometry, strange attractors, multistability, fractal basin boundaries, and Wada basins of attraction. We also develop slow-fast dynamical systems theory for discrete-time systems, with a specific application to modeling the spiking and bursting behavior emerging from the electrophysiology of biological neurons. Our research mainly focuses on two simple low-dimensional slow-fast Rulkov maps, which model both non-chaotic and chaotic spiking-bursting neuronal behavior. We begin by exploring the maps' individual dynamics and parameter spaces, performing bifurcation analyses, describing and quantifying their chaotic dynamics, and modeling an injection of current into them. Then, by putting these neurons into different physical arrangements and coupling them with a flow of current, we find that complex dynamics and geometries emerge from the existence of multistability and sensitivity to initial conditions in higher-dimensional state space. We then analyze the complexity and fractalization of these coupled neuron systems' attractors and basin boundaries using our mathematical and computational methods. This paper begins with a conversational introduction to the geometry of chaos, then integrates mathematics, physics, neurobiology, computational modeling, and electrochemistry to present original research that provides a novel perspective on how types of geometrical sensitivity to initial conditions appear in discrete-time neuron systems.
Chaos Modeling: A Comparison of Classical and Quantum Reservoir Computer Capabilities
Arjun Bhat, Kanjonavo Sabud
Dynamic systems have always been an integral part of our world. Computational models are being explored for their potential of modeling dynamic systems ranging from storm development and wildfire behavior to natural language understanding and stock market prediction. The development of an RNN subfield known as Reservoir Computing (RC) has received wide attention and is observed to be well suited for handling dynamic system forecasting. Likewise, quantum machine learning has also been seen to increase model capabilities. In this project, we researched on combining the two by implementing a Quantum Reservoir Computer (QRC) for modeling a benchmark dynamic system -- the Lorenz 1963. Furthermore, we used a standard Artificial Neural Network (ANN) as a control model. Our results showed that Classical RC with forecast horizons averaging 90% in train lengths of 9000 timesteps greatly outperformed the ANN model which averaged 30%. However, our QRC model completely failed to model the Lorenz 63 system and achieved a 0% forecast horizon. As research on QRC and its dynamic systems modeling capabilities is still in its infancy, it is highly likely that our implementation of QRC was not prepared to model a system as complex as Lorenz 63. Future work, therefore, includes improving our implementation of QRC. It would also be interesting to explore the intersection of graph theory and networks within the reservoir of a reservoir computer.
Development of a Scalable Silicon Photonic On-Chip Memory Architecture
Sathvik Redrouthu, Pranav Vadde, Pranav Velleleth
Over the past decade, there has been a dramatic increase in the parameter count of neural networks, driven by advances in machine learning algorithms, hardware, and data availability. This increase has enabled significant improvements in performance on a wide range of tasks, from image classification to natural language processing. In 2012, the state-of-the-art image classification model, AlexNet, had only 61 million parameters. By 2015, the VGG-19 model had 143 million parameters, and by 2016, the ResNet-152 model had 60 million parameters. In 2018, the DenseNet-264 model had 36.4 million parameters, while the EfficientNet-B7 model, released in 2019, had 66 million parameters. The parameter counts of natural language models have also increased significantly. In 2015, the state-of-the-art language model had only 5 million parameters. By 2018, the OpenAI GPT-2 model had 1.5 billion parameters, and by 2020, the GPT-3 model had 175 billion parameters. This increase in parameter count has raised concerns about the computational cost and environmental impact of training and running these models. For example, training the GPT-3 model on a single accelerator can consume up to 1.2 GWh of electricity. To address these concerns, in addition to exploring various techniques for reducing the neural network parameter counts while maintaining performance, we have developed a set of silicon photonic accelerators with significantly higher speed and energy ratings for inference processes. Despite these advantages, these accelerators don’t have the capabilities to efficiently execute larger models, as data must be converted to electrical signals for traditional intermediate memory storage. Thus, we explore nonvolatile optical memory, with the goal of removing these intermediate conversions and improving overall performance. We begin by evaluating the effectiveness of on-chip silicon photonic memory architectures; notability, those using Phase Change Materials. We then turn to the original free-space experiments within nonvolatile optical memory and go on to design an experiment taking these to the silicon photonic domain. We address introduced issues like on-chip crosstalk through novel innovations to introduce scalability with optical memory cells, which was previously not possible. We finish by evaluating our on-chip memory system against conventional systems and other silicon photonic architectures in literature in “storage time”, “scalability”, “storage capacity”, and “read/write time.”
Simulating Self-Gravitating Dark Matter with Variational Quantum Computing
Abhinav Angirekula, Dhruv Anurag, Rohan Kompella
Galaxies have far too little observable matter for them to be gravitationally bound. Thus, there must be a theoretical, unobservable from of matter that makes up this missing mass: dark matter. Dark matter does not interact with light and we can only see its effects through gravitational lensing. The current leading candidates for dark matter are weakly interacting massive particles (WIMPs), primordial black holes, and axions. Accurate dark matter simulations are vital for researching their true nature, as scientists can compare simulations to their observations. If there are discrepancies between the two, there might be an undiscovered property or interaction that dark matter has. While dark matter simulation appears to be incredibly complicated, it really just boils down to time evolving a system using differential equations. In our case, the Schrödinger-Poisson equations are a system of non-linear differential equations that govern the evolution of several types of dark matter models: from fuzzy dark matter to standard cold dark matter. Mocz and Szasz were able to solve the Schrödinger-Poisson equations using a classical spectral method, and used a variational quantum algorithm outlined by Lubasch et. al. and were able to successfully model dark matter. However, instead of running their quantum algorithm on a quantum computer, they ran it on a simulation of a quantum computer on a classical machine. For our project, we want to replicate the results of Mocz and Szasz and implement both their classical and quantum algorithms. We want to run their quantum algorithm on an actual quantum computer (e.g. IBM), and the number of qubits used in it is under the maximum limit of quantum computers we could use so that we can compare the efficiency of each algorithm.
Presentation Video (FCPS Only)
Exploring Quantum Finance
Thomas Winston
Quantum Computing is unquestionably the future. There are 2 general architectures behind any quantum computer, quantum annealing or a gate based approach. Quantum annealing is a way to solve an incredible spectrum of optimization problems. Even better, there is only 1 company actually generating a profit using quantum computing, and that is DWAVE, a quantum annealing company. My project specifically focuses on using quantum annealing to solve the large scale optimization problem that is the stock market, through a mathematical framework called Modern Portfolio Theory, which breaks down portfolio optimization into a basic weight of risk vs reward.
Tracking Microplastics with Quantum Machine Learning and Dynamics
Anirudh Mantha
Plastic pollution in the ocean has become a major concern in recent years. Approximately 400 million tons of plastic waste are generated annually, and when this plastic enters the ocean, it poses a threat to marine life. Microplastics, which are small plastic particles that break off from larger plastic clumps, are particularly hazardous. These microplastics are difficult to track, as current methods rely on detecting surfactants, chemicals that reduce the surface tension between two liquids. However, studies have shown that surfactants are often associated with microplastics, this takeaway has led to a lot of research being done tracking microplastics by measuring the surface tension of the water and seeing the surfactant concentration. If there are surfactants, they assume that there are microplastics there, however this is not an accurate measure of figuring out this. To address this issue, it is necessary to develop a more effective method for tracking microplastics. One potential solution is to use a combination of remote sensing and mathematical analysis of ocean current models. Machine learning could be used to locate the initial plastics, while a mathematical model could be used to predict their future location of when they branch off. There are many variables to consider when modeling ocean currents, such as wind, water density, gravity, storms, and biomes. By successfully integrating these two approaches, we may be able to accurately predict the location of microplastics and aid in their removal.
Implementing Quantum Key Distribution using Photonics
Katherine Jimenez, Alec Riso, Karthik Thyagarajan, Geoffrey Whiting
Quantum Key Distribution (QKD) is a revolutionary cryptographic protocol that leverages the principles of quantum mechanics to establish secure communication channels. Unlike classical encryption methods, QKD relies on the fundamental properties of quantum states and ensures that any attempt to eavesdrop on the key exchange introduces detectable disturbances, thus guaranteeing the security of transmitted information. Currently, the most viable way to implement QKD for communication is via photonics — namely, using phase-preserving long-distance optical fibers. The objective for our project is to implement QKD in photonics in a laboratory setting which will help advance the protocol's robustness and feasibility for practical use. Furthermore, we aim to investigate the difficulty of noise in implementing the algorithm over long distances.
Exploring a Model of Quantum Cognition
Christopher Qian
Math is useful because it creates and derivates quantifiable expressions to explain events that occur in reality. In this case, the amalgamation of Quantum theory into Cognitive Science attempts to patch holes inside the latter, such as the Disjunction Effect illustrated in the Prisoner's Dilemma, as a new branch known as Quantum Cognition. However, its novelty has yet to be tested with quantum computational servers, relying on the NumPy library to calculate results manually. This project will achieve three objectives: use Quantum Cognition to model the Iowa Gambling Task, utilize Qiskit (a Quantum Computing library) and Reinforcement Learning to model a Quantum Cognition agent, and prove quantum superiority by pitting the Quantum agent against a standard Q-learning agent.
[Research Poster] https://drive.google.com/file/d/1JDXNdUPGvrKEP6wHfDqP0aO5gIWap-nf/view?usp=sharing
[Research Presentation Slides] https://drive.google.com/file/d/16Ay5Uu2Si0UQjJBX31bSIgTNRPlcKP0b/view?usp=sharing
[Presentation Video] https://drive.google.com/file/d/1aEDAaoou7RahUJGFNSTgy9t6RLFF7oPT/view?usp=sharing
Using Quantum ML and Analyzing Electro-Optical vs SAR Image Classification Through Polarizati
Naisha Patel
Intelligence agencies aim to find the most efficient methods to combat adversaries. The efficiency of image processing from satellite images is essential to ensure information about the world is being recorded. Two types of images, SAR and Electro-Optical Images, have different purposes, advantages, and disadvantages. SAR Images are reliable during any time of day across any weather type. Electro-Optical Images are dependent on the sun's presence, however, create more detailed images. Both types of images may not be valued the same amount depending on how well they can classify certain images in certain areas. Using Quantum ML to create my classification model, and using a dataset of civilian vehicles that have been modified to account for polarizations, noise, and terrain, I aim to investigate which type of image is more beneficial for intelligence agencies.
Using Physics-Informed Neural Networks to Model the 2D Heat Equation
Mihika Dusad, Michelle Kang, Emi Zhang, Laura Zhang
Partial differential equations (PDEs) are essential tools in various fields, but lack closed-form solutions, requiring approximation methods. Recent advancements like the Adomian domain decomposition method and the optimal auxiliary function method (OAFM) have improved PDE solution techniques, with OAFM offering linearization and efficient convergence in one interaction. Alongside mathematical approaches, computational tools like physics-informed neural networks (PINNs) have emerged, leveraging physics-based error terms for accuracy. Our project aims to integrate mathematical and computational methodologies to model the 2D heat equation, enhancing PINNs with classical analysis techniques and real-world data.
Exploring Fischer Koch S as a Photonic Crystal
Srimaye Peddinti
Photonic crystals are nanomaterials with periodic refractive indices. Their periodic properties create photonic band gaps, which prohibit certain wavelengths from propagating through the material, such that it iridesces a specific range of colors when exposed to light. Recently, triply periodic minimal surfaces (TPMS) have been targeted as viable candidates for photonic crystal applications. This research aims to explore a specific TPMS, the Fischer Koch S structure, and establish its candidacy as a photonic crystal. The main goal of this research is to further the field of photonics and potentially create more responsive photonic devices.
Teaching Bell's Theorem and Entanglement Through Online Simulations
Krish Malik
The recent Nobel Prize in physics was awarded to Alain Aspect, John Clauser, and Anton Zeilinger for their experiments with entangled photons, which showed a violation of Bell inequalities. One method of describing entanglement is quantum coin flipping, an idea that allows us to explore the connections to classical coin flipping. It uses qubit measurement gates to highlight the interconnectivity of entangled particles and the mathematics behind these connections. Our project will attempt to show violations of Bell inequalities in an online simulation. Most experiments have dealt with hands-on entanglement methods, such as the one done by Aspect, Clauser, and Zeilinger. The simulations will require newer methods of describing entanglement, such as developing a quantum coin-flipping platform through Qiskit–a computer package that permits the creation of various measurement gates. Our project will be able to use classical computing methods to show the quantum entanglement of particles in violation of Bell inequalities.
Building and Entangled Photon Pair Source
Steven Lu, Lucas Marschoun
Photon pair sources are fundamental for quantum computing, optics, instrumentation, and experimentation, with numerous applications such as imaging cells and quantum key distribution. Spontaneous Parametric Down-Conversion (SPDC) is the most common and mature process used today to generate entangled photon pairs. It uses a non-linear crystal to produce spin entangled photon pairs (signal and idler) from a photon beam (pump). Our goal is to build a source that exhibits Type-I SPDC and use it to experimentally test Bell's inequality. Bell's inequalities are a set of inequalities that when violated, contradict the local hidden variable theory of quantum mechanics. In our case, we would aim to find a discrepancy between the predicted and actual quantities of photons produced from our emitter that pass through a pair of optical polarizers. In our project we work through setting up the various components of the photon emitter and aligning it.
Constrained Deep Learning Methods for Optimization of Autonomous Systems
Brian Zhou
The last few decades have led to the rise of research focused on propulsion and control systems for bio-inspired unmanned underwater vehicles (UUVs), which provide more maneuverable alternatives to traditional UUVs in underwater missions. Recent work has explored the use of time-series neural network surrogate models to predict thrust and power from vehicle design and fin kinematics. We develop a search-based inverse model that leverages kinematics-to-thrust and kinematics-to-power neural network models for control system design. Our inverse model finds a set of fin kinematics with the multi-objective goal of reaching a target thrust under power constraints while creating a smooth kinematics transition between flapping cycles. We demonstrate how a control system integrating this inverse model can make online, cycle-to-cycle adjustments to prioritize different system objectives, with improvements in increasing thrust generation or reducing power consumption of any given movement upwards of 0.5 N and 3.0 W in a range of 2.2 N and 9.0 W. As propulsive efficiency is of utmost importance for flapping-fin UUVs in order to extend their range and endurance for essential operations but lacks prior research, we develop a non-dimensional figure of merit (FOM), derived from measures of propulsive efficiency, that is able to evaluate different fin designs and kinematics, and allow for comparison with other bio-inspired platforms. We use the developed FOM to analyze optimal gaits and compare the performance between different fin materials, providing a better understanding of how fin materials affect thrust generation and propulsive efficiency and allowing us to inform control systems and weight for efficiency on the developed inverse gait-selector model.
Quantum Holography
Lindsay Hwang, Aditya Sengar
Holography is a method of representing a 3D object in a 2D way. Classical holography exists in many forms, from transmission to rainbow. However, quantum holography is much rarer, in which it uses quantum entanglement, where the state of one particle becomes inextricably linked with another's, regardless of spatial separation, to create a hologram. In our experiment, we first created a classical transmission hologram using the LitiHolo hologram kit, imprinting the object's information on a holographic plate. We then created an experimental setup for our quantum hologram, with a laser pointed at a ppKTP crystal that would perform spontaneous parametric down conversion (SPDC). This allows for the two beams, which usually recombine in classical holography, to never combine following their split while creating the hologram. We believe that there is a promising future for this project later on, as we have established both a setup and a connection with Sebastian Töpfer, the lead researcher in the quantum holography paper that our experiment is based on. As for this project's implications, as quantum holography continues to evolve, it promises transformative breakthroughs in the biomedical field in internal imaging and holography overall.
Using Quantum Agents to Model Simple Economies
Daniel Sprintson, Brian Zhou
Agent-based modeling can study any social behavior and decision-making process, as agents make decisions based on an internal logic network and try to imitate human behavior. Proponents say that models based on agent interactions can inspire insight into policy and predict aggregate human behavior. Detractors concern themselves with the applicability of these results and whether such agents can fundamentally capture the nuance of human behavior. The solution lies in additional scale and complexity, often expensive to simulate and impossible with current computational methods for simulation. In this paper, we explore a new approach to agent-based modeling incorporating the properties of quantum mechanics and quantum computing. We build quantum adaptive long-term learning agents to measure the influence of external stimuli on a simple business structure comprised of those agents and evaluate the performance benefits from a quantum approach, comparing results with conventional Bayesian and Computational agents.
Generative Modeling with the Yukawa Potential
Pranav Kuppa, Max Wang, David Wei, Zani Xu
Diffusion models, the current state-of-the-art in image generation, have led the charge in generative modeling for several years. We provide a theoretical derivation and implementation of a new generative model based on the Yukawa potential. This model is the first of a new generation of models with nonzero birth rates, offering for more creative, and in some cases, accurate generation. The nonzero birth rate is a new variable in the generative process, offering great implications in applications without a constant number of particles. We focused on image generation as a proof of concept.
Engineering Artificial Muscles
Siri Duddella, Sankarshanaram Vempati
Long title: SPARK’EM: Simulating (nanocomposite) Polymers via Ab-Initio methods to realize to Realize high-(K)apacitance Elastomeric Artificial Muscles
Hydraulically amplified self-healing electrostatic actuators, or HASELs, are an emerging class of dielectric elastomeric actuators that mimic the behavior of human muscular systems. By permeating an electric field across a dielectric elastomeric “pouch” filled with dielectric oil, hydraulic pressure is developed and the actuator can support loads. However, generating this electrostatic potential requires a high voltage requirement (~25 kV) and a flexible yet robust elastomer, making implementing this useful soft-robotic actuator difficult in traditional low-voltage DC systems. This problem can be rectified by increasing the dielectric constant (k) of the elastomer via the integration of high-k nanomaterials with durable low-k thin-film elastomers, allowing the actuator to function at significantly lower voltages for thousands of cycles without breakdown. To develop a novel material for this task, we propose a computational pipeline utilizing quantum descriptors (density functional theory (DFT) calculated energy and polarizability) in conjunction with copolymer molecular dynamics to discover and model candidate materials. Piping this data into LAMMPS, a soft material and robotics simulator, we evaluate elastomer candidates' efficacy in robotic applications and select a final material, P(VDF-TrFE-CTFE) with Ti3C2Tx MXene, for thin-film synthesis and actuator integration. The result was an actuator that could perform linear compressions at only 1.1kV, a 96% decrease from 25kV in our original model. We also propose future research and testing of our model by tuning the ReaxFF parameters we used to match the QM model better, integrating more rigorous testing methods, and constructing a full-scale actuator.
Detecting and Identifying Floor Vibrations through Interferometers
Shashank Cheruvu
Interferometers are laser-based investigative tools that are highly sensitive to vibrations. Because of this, when placed on the floor, they can detect the small differences between the walking styles of different people, meaning that, if attached to a sufficiently trained neural network, we can discern the identity of a person who is walking without needing audio or video. This could be applied in the future as a surveillance tool, or more!
An Exploration into Dynamic Spin Chemistry: Improving NMR Resolution
Mohib Ahmed
Nuclear magnetic resonance(NMR) is a powerful technique to resolve structural information about molecules. Of particular interest is NMR of proteins, where understanding the structural properties takes an especial importance. However, NMR techniques are limited by resolution. In proteins with potential molar masses in the multiple thousands, resolving NMR data can be difficult. This is where Dynamic Nuclear Polarization (DNP) comes in. DNP is a well characterized phenomena where the nuclear magnetic resonance of electrons is enhanced by external polarization of those electrons. In Kaushik et al., 2016, this was achieved on proteins with spin-labeling with Gadolinium (III) and Manganese (II) complexes. In Biedenbänder et. al., 2022, it was shown that dynamic nuclear polarization is a process capable of working on improving resolution of proteins. NV Diamonds have the benefit of effecting general dynamic nuclear polarization without chemically altering the protein. The established method of DNP enhancing proteins, spin labeling, is more targeted, however it does involve chemically altering the proteins. Hence, this project will investigate the effects of NV diamonds on DNP in methanol, with the hopes of justifying its usage in specific protein NMR applications.
Using Quantum Control to Achieve Spectral Super-resolution
Taohan Lin
Quantum systems have long shown promise for detecting low magnitude environmental effects such as electric and magnetic fields. We consider the use of quantum systems to detect time-varying electromagnetic signals. Expanding on a recent application of quantum sensing obtaining spatial super-resolution in optical imaging, we aim to apply super-resolution techniques to resolve adjacent spectral peaks in frequency analysis.
Previous research showed that frequency super-resolution is possible, and discovered one control sequence that achieves it. The optical super-resolution techniques allow us to identify optimal conditions for the applied control to the quantum sensor. In this project, we numerically optimize control sequences for super-resolution, and compare the optimized sensing protocol with other methods for frequency analysis. We consider the family of two π-pulse sequences with specific evolution times and evaluate sequences by the Fisher information the corresponding sensing protocol provides. In this family, we find that the Carr-Purcell-Meiboom-Gill sequence (CPMG), a commonly used sequence in quantum sensing protocols, is optimal for resolving the frequency separation between spectral peaks. We show that the method using CPMG theoretically allows for peak resolution even as the frequency separation approaches zero, and we approximate protocol error bounds under experimental conditions. We identify conditions when the super-resolution protocol outperforms the classical discrete Fourier transform and quantum noise spectroscopy methods when data collection time is held constant. The protocol developed is applicable in current-day nitrogen vacancy diamond quantum sensors and can offer advances for frequency peak resolution within nuclear magnetic resonance spectroscopy, magnetometry, and orthogonal frequency division multiplexing.
Analyzing Complexity in Variational Quantum Algorithms
Taohan Lin
Variational quantum algorithms are a class of algorithms which make use of both classical and quantum techniques to solve a problem. The quantum approximation optimization algorithm (QAOA) is one such algorithm which focuses on solving combinatorial optimization problems. In this study, we numerically approximate the computational complexity of the QAOA for the Ising model and calculate complexity bounds. We create a simulator to calculate the ground state and ground state energy of the TFIM for arbitrary system size and coupling coefficient. At the phase transition, we observe the existence of a critical layer count for QAOA after which error in the calculated ground state energy drops drastically. We also observe that this critical layer count increases roughly linearly with system size. When the system size is held constant, We find that error in the ground state energy is maximized near the phase transition prior to achieving the critical layer count, and minimized after achieving the critical layer count. The greatest deviation in error before and after the critical layer count is at the phase transition. While the study is limited by the small system sizes we studied, the results could still be quite useful for understanding and optimizing the QAOA algorithm in the future.
Optimization of Traffic Lights Using Quantum Behaving Particles
William Kerr, Benjamin Rubin
Particle swarm optimization (PSO) is an optimization algorithm that mimics the flocking behavior of birds, ants, and other animals in order to discover the minimum of a cost function. We implemented and improved quantum particle swarm optimization (QPSO), a variation of PSO representing each particle as a quantum wave-function. We tested various potential functions to determine the optimal function for our purposes. This gave both a more efficient and more reliable algorithm. For our project we applied this algorithm to minimize wait times at local traffic lights, inputting data into the traffic simulator SUMO and using the lengths of light cycles as the variables.
Implementing QKD-Encrypted Hybrid Quantum Machine Learning for Dementia Detection
Ryan Kim
Magnetic resonance imaging (MRI) is a common technique to scan brains for strokes, tumors, and other abnormalities that cause forms of dementia. However, correctly diagnosing forms of dementia from MRIs is difficult, leading to misdiagnoses, an issue neural networks can rectify. The performance of these neural networks has been shown to be improved by applying quantum algorithms. This proposed approach takes advantage of a classical convolutional neural network (CNN) to extract features, and then uses a quantum support vector machine (QSVM) to classify a given instance. This study hopes to improve the accuracy and efficiency of classical CNNs. With hospitals beginning to adopt machine learning applications for biomedical image detection, this proposed architecture could improve accuracies and prevent more misdiagnoses. Additionally, to address privacy concerns over sensitive patient data, quantum key distribution (QKD) for image encryption is employed before classification. Furthermore, the proposed architecture is flexible and can be used for transfer-learning tasks, saving time and resources.
BlochAR: Developing a Manipulative Bloch Sphere Implemented as an Augmented Reality (AR) Application
Alina Chen, Rishabh Chhabra
The Bloch sphere is a geometrical representation of the pure state space of a two-level quantum mechanical system. The realistic visualization of quantum states through the Bloch sphere will provide an intuitive understanding of complex quantum phenomena, which can often seem abstract and counterintuitive. Our project aims to revolutionize educational methods in quantum physics and mechanics. To do this, we want to implement the Bloch sphere as an AR application through the game engine Unity and the HoloLens goggles. Our application offers a more interactive and engaging way to explore the Bloch sphere that reaches far beyond traditional teaching methods.
Building a Magneto-Optical Trap
Siddhita Krishnan, Ananya Pamal
A Magneto-Optical Trap (MOT) uses lasers to slow ions down to extremely low speeds in a millikelvin temperature range so their quantum properties can be studied. Existing MOTs have thus far been used to cool and slow down twenty elements, with the creators receiving the Nobel Prize in Physics in 1997. Using the MOT to study atoms can reveal insights into energy generation and speeding up quantum computing. Rubidium MOTs in particular are some of the most common, and we will be creating one to help supplement future projects into quantum computing and quantum energy.
On Optimizing Quantum Cosmological Simulations of Self-Gravitating Dark Matter
David Cao, Ronit Kapur, Rishabh Prabhu
Today, visible matter makes up less than 5% of the total mass in the universe. The rest of the mass can be largely attributed to an invisible matter known as dark matter. Although dark matter is invisible, it is integral to the structure of the universe, as galaxies, clusters, and other deep sky objects are primarily composed of dark matter. Because dark matter is so important to understanding the inner workings of the universe, it is essential to be able to model it. Traditional classical simulations struggle to do this, as modeling dark matter requires extensive computational time and space—it requires approximating solutions to the nonlinear differential equations that describe dark matter dynamics. These differential equations—the Vlasov-Poisson (VP) equations—have been formulated to approximate the time evolution of self-gravitating dark matter.. Previous studies find that, in the cases of cold dark matter and general self-gravitating dark matter respectively, the VP equations can be approximated by the Schrödinger-Poisson (SP) equations. Not only are the SP equations more applicable to quantum simulations—as they directly evolve a wave function—but they also exhibit smoother solutions without discontinuities or singularities. We apply a hybrid classical-quantum algorithm, specifically variational quantum computing (VQC), an approach analogous to machine learning. In doing so, one addresses the computing limitations of purely classically simulating dark matter. We build on previous work by simulating cosmological dark matter in 2D space and account for the exponential build-up of error over compounding time intervals.
Partial Least-Squares Quantum Optimization
Parth Gupta
Previously, I used a partial least squares (PLS) implementation to predict March Madness brackets, called the Massey Method. I learned that, although accurate, it was a very slow algorithm. My ultimate goal for this project is to optimize the classical implementation of a least squares algorithm using the Quantum Approximate Optimization Algorithm (QAOA). The classical least squares implementation would take too long for a computer to diagnose the patient. Instead, the doctor would have to make an educated prediction and hope their treatment would sufficiently aid the patient. With a QAOA implementation of PLS, the doctor would receive a quick and accurate diagnosis, enabling them to administer treatment quickly and correctly, potentially saving the patient’s life. Of course, this is just one example of the power of this algorithm and can be applied to many other scenarios.
Transferring Quantum Populations Using STIRAP
Laura Guo, Danielle Riekse, Ishara Shanmugasundaram
One of the biggest hurdles in modern-day quantum physics is figuring out how to transfer quantum populations from one state to another. Stimulated Raman Adiabatic Passage (STIRAP) aims to solve this problem by creating an efficient, active method of transfer. STIRAP relies on three-state pulsed stimulated Raman scattering (SRS). Raman scattering induces a change in the vibrational state of the atom due to absorption of energy, and this scattering can be stimulated using a pump and stroke photon. The pump photon excites the atom to a virtual state while the stoke photon relaxes it to a different state than the original. The adiabatic nature of STIRAP passages allows it to not get affected by external perturbation. Our goal is to create a STIRAP model with 75% efficiency of transfer and ideally implement it in superconducting quantum circuits.