Search Talks

The introduction of neural networks with deep architecture has led to a revolution, giving rise to a new wave of technologies empowering our modern society. Although data science has been the main focus, the idea of generic algorithms which automatically extract features and representations from raw data is quite general and applicable in multiple scenarios. Motivated by the effectiveness of deep learning algorithms in revealing complex patterns and structures underlying data, we are interested in exploiting such tool in the context of many-body physics. I will first introduce the Boltzmann Machine, a stochastic neural network that has been extensively used in the layers of deep architectures. I will describe how such network can be used for modelling thermodynamic observables for physical systems in thermal equilibrium, and show that it can faithfully reproduce observables for the 2 dimensional Ising model. Finally, I will discuss how to adapt the same network for implementing the classical computation required to perform quantum error correction in the 2D toric code.

An increase in the efficiency of sampling from Boltzmann distributions would have a significant impact in deep learning and other machine learning applications. Recently, quantum annealers have been proposed as a potential candidate to speed up this task, but several limitations still bar these state-of-the-art technologies from being used effectively. One of the main limitations is that, while the device may indeed sample from a Boltzmann-like distribution, quantum dynamical arguments suggests it will do so with an instance-dependent effective temperature, different from the physical temperature of the device. Unless this unknown temperature can be unveiled, it might not be possible to effectively use a quantum annealer for Boltzmann sampling. In this talk, we present a strategy to overcome this challenge with a simple effective-temperature estimation algorithm. We provide a systematic study assessing the impact of the effective temperatures in the learning of a kind of restricted Boltzmann machine embedded on quantum hardware, which can serve as a building block for deep learning architectures. We also provide a comparison to k-step contrastive divergence (CD-k) with k up to 100. Although assuming a suitable fixed effective temperature also allows to outperform one step contrastive divergence (CD-1), only when using an instance-dependent effective temperature we find a performance close to that of CD-100 for the case studied here. We discuss generalizations of the algorithm to other more expressive generative models, beyond restricted Boltzmann machines.

Machine learning is a rapidly growing field in computer science with applications in computer vision, voice recognition, medical diagnosis, spam filtering, search engines, etc. In this presentation, I will introduce a new machine learning approach based on quantum Boltzmann distribution of a transverse-field Ising Model. Due to the non-commutative nature of quantum mechanics, the training process of the Quantum Boltzmann Machine (QBM) can become nontrivial.  I will show how to circumvent this problem by introducing bounds on the quantum probabilities. This allows training the QBM efficiently by sampling. I will then show examples of QBM training with and without the bound, using exact diagonalization, and compare the results with classical Boltzmann training. Finally, after a brief introduction to D-Wave quantum annealing processors, I will discuss the possibility of using such processors for QBM training and application.

We propose a family of models with an exponential number of parameters, but which are approximated by a tensor network. Tensor networks are used to represent quantum wavefunctions, and powerful methods for optimizing them can be extended to machine learning applications as well. We use a matrix product state to classify images, and find that a surprisingly small bond dimension yields state-of-the-art results. Tensor networks offer many advantages for machine learning, such as better scaling for existing machine learning approaches and the ability to adapt hyperparameters during training. We will also propose a generative interpretation of the trained models.

In the first part of this talk, I will focus on the physics of deep learning, a popular subfield of machine learning where recent performance on tasks such as visual object recognition rivals human performance. I present work relating greedy training of deep belief networks to a form of variational real-space renormalization. This connection may help explain how deep networks automatically learn relevant features from data and extract independent factors of variation. Next, I turn to the information bottleneck (IB), an information theoretic approach to clustering and compression of relevant information that has been suggested as a framework for deep learning. I present a new variant of IB called the Deterministic Information Bottleneck, arguing that it better captures the notion of compression while retaining relevant information.

Quantum control is valuable for various quantum technologies such as high-fidelity gates for universal quantum computing, adaptive quantum-enhanced metrology, and ultra-cold atom manipulation. Although supervised machine learning and reinforcement learning are widely used for optimizing control parameters in classical systems, quantum control for parameter optimization is mainly pursued via gradient-based greedy algorithms. Although the quantum fitness landscape is often compatible for greedy algorithms, sometimes greedy algorithms yield poor results, especially for large-dimensional quantum systems. We employ differential evolution algorithms to circumvent the stagnation problem of non-convex optimization, and we average over the objective function to improve quantum control fidelity for noisy systems. To reduce computational cost, we introduce heuristics for early termination of runs and for adaptive selection of search subspaces. Our implementation is massively parallel and vectorized to reduce run time even further. We demonstrate our methods with two examples, namely quantum phase estimation and quantum gate design, for which we achieve superior fidelity and scalability than obtained using greedy algorithms.

Quantum machine learning algorithms usually translate a machine learning methods into an algorithm that can exploit the advantages of quantum information processing. One approach is to tackle methods that rely on matrix inversion with the quantum linear system of equations routine. We give such a quantum algorithm based on unregularised linear regression. Opposed to closely related work from Wiebe, Braun and Lloyd [PRL 109 (2012)] our scheme focuses on a classification task and uses a different combination of core routines that allows us to process non-sparse inputs, and significantly improves the dependence on the condition number. The second part of the talk presents an idea that transcends the reproduction of classical results. Instead of considering a single trained classifier, practicioners often use ensembles of models to make predictions more robust and accurate. Under certain conditions, having infinite ensembles can lead to good results. We introduce a quantum sampling scheme that uses the parallelism inherent to a quantum computer in order to sample from 'exponentially large' ensembles that are not explicitely trained.

Supervised Machine Learning is one of the key problems that arises in modern big data tasks. In this talk, I will first describe several different classical algorithmic paradigms for classification and then contrast them with quantum algorithmic constructs. In particular, we will look at classical methods such as the nearest neighbor rule, optimization based algorithms (e.g. SVMs), Bayesian inference based techniques (e.g. Bayes point machine) and provide a unifying framework so that we can get a deeper understanding about the quantum versions of the methods.

Tensor networks have been very successful for approximating quantum states that would otherwise require exponentially many parameters.

I will discuss how a similar compression can be achieved in models used to machine learn data, such as sets of images, by representing the fitting parameters as a tensor network. The resulting model achieves state-of-the-art performance on standard classification tasks. I will discuss implications for machine learning research, exploring which insights from physics could be imported into this field.

In this talk I address the problem of simultaneously inferring unknown quantum states and unknown quantum measurements from empirical data. This task goes beyond state tomography because we are not assuming anything about the measurement devices. I am going to talk about the time and sample complexity of the inference of states and measurements, and I am going to talk about the robustness of the minimal Hilbert space dimension. Moreover, I will describe a simple heuristic algorithm (alternating optimization) to fit states and measurements to empirical data. For this algorithm the dataset does not need to be quantum. Hence, the proposed algorithm enables us to interpret general datasets from a quantum perspective. By analyzing movie ratings, we demonstrate the power of quantum models in the context of item recommendation which is a key discipline in machine learning. We observe that quantum models can compete with state-of-the-art algorithms for item recommendation. Based on joint work with Aram Harrow. Relevant preprints: arXiv:1412.7437 and arXiv:1510.02800.