Peter Wittek

Scalable agent-based modelling with cloud HPC resources for social simulations

New concepts like agent-based modelling are providing social scientists with new tools, more suited to their background than other simulation techniques. The success of this new trend will be strongly related to the existence of simulation tools capable of fulfilling the needs of these disciplines. Given the computational requirement of realistic agent-based models, high-performance computing infrastructure is often necessary to perform the calculations. At present, such resources are unlikely to be available to humanities researchers. Having developed Pandora, an open-source framework designed to create and execute large-scale social simulations in high-performance computing environments, this work presents an evaluation of the impact of cloud computing within this context. We find that the constraints of the cloud environment do not have a significant impact on the generic pattern of execution, providing a cost-effective solution for social scientists.

Accelerating text mining workloads in a MapReduce-based distributed GPU environment

Scientific computations have been using GPU-enabled computers successfully, often relying on distributed nodes to overcome the limitations of device memory. Only a handful of text mining applications benefit from such infrastructure. Since the initial steps of text mining are typically data intensive, and the ease of deployment of algorithms is an important factor in developing advanced applications, we introduce a flexible, distributed, MapReduce-based text mining workflow that performs I/O-bound operations on CPUs with industry-standard tools and then runs compute-bound operations on GPUs which are optimized to ensure coalesced memory access and effective use of shared memory. We have performed extensive tests of our algorithms on a cluster of eight nodes with two NVidia Tesla M2050s attached to each, and we achieve considerable speedups for random projection and self-organizing maps.

Compactly Supported Basis Functions as Support Vector Kernels for Classification

Wavelet kernels have been introduced for both support vector regression and classification. Most of these wavelet kernels do not use the inner product of the embedding space, but use wavelets in a similar fashion to radial basis function kernels. Wavelet analysis is typically carried out on data with a temporal or spatial relation between consecutive data points. We argue that it is possible to order the features of a general data set so that consecutive features are statistically related to each other, thus enabling us to interpret the vector representation of an object as a series of equally or randomly spaced observations of a hypothetical continuous signal. By approximating the signal with compactly supported basis functions and employing the inner product of the embedding L2 space, we gain a new family of wavelet kernels. Empirical results show a clear advantage in favor of these kernels.

Learning in quantum control: High-dimensional global optimization for noisy quantum dynamics

Abstract 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 with 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. We improve quantum control fidelity for noisy system by averaging over the objective function. 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.

Unbounded randomness certification using sequences of measurements

Unpredictability, or randomness, of the outcomes of measurements made on an entangled state can be certified provided that the statistics violate a Bell inequality. In the standard Bell scenario where each party performs a single measurement on its share of the system, only a finite amount of randomness, of at most 4logd bits, can be certified from a pair of entangled particles of dimension d. Our work shows that this fundamental limitation can be overcome using sequences of (nonprojective) measurements on the same system. More precisely, we prove that one can certify any amount of random bits from a pair of qubits in a pure state as the resource, even if it is arbitrarily weakly entangled. In addition, this certification is achieved by near-maximal violation of a particular Bell inequality for each measurement in the sequence.

High-performance dynamic quantum clustering on graphics processors

Clustering methods in machine learning may benefit from borrowing metaphors from physics. Dynamic quantum clustering associates a Gaussian wave packet with the multidimensional data points and regards them as eigenfunctions of the Schrodinger equation. The clustering structure emerges by letting the system evolve and the visual nature of the algorithm has been shown to be useful in a range of applications. Furthermore, the method only uses matrix operations, which readily lend themselves to parallelization. In this paper, we develop an implementation on graphics hardware and investigate how this approach can accelerate the computations. We achieve a speedup of up to two magnitudes over a multicore CPU implementation, which proves that quantum-like methods and acceleration by graphics processing units have a great relevance to machine learning.

Improving Text Classification by a Sense Spectrum Approach to Term Expansion

Experimenting with different mathematical objects for text representation is an important step of building text classification models. In order to be efficient, such objects of a formal model, like vectors, have to reasonably reproduce language-related phenomena such as word meaning inherent in index terms. We introduce an algorithm for sense-based semantic ordering of index terms which approximates Cruses description of a sense spectrum. Following semantic ordering, text classification by support vector machines can benefit from semantic smoothing kernels that regard semantic relations among index terms while computing document similarity. Adding expansion terms to the vector representation can also improve effectiveness. This paper proposes a new kernel which discounts less important expansion terms based on lexical relatedness.

Somoclu : An Efficient Parallel Library for Self-Organizing Maps

Somoclu is a C++ tool for training self-organizing maps on large data sets using a high-performance cluster. It builds on MPI for distributing the workload across the nodes of the cluster. It is also able to boost training by using CUDA if graphics processing units are available. A sparse kernel is included, which is useful for high-dimensional but sparse data, such as the vector spaces common in text mining workflows. The code is released under GNU GPLv3 licence.

Monitoring term drift based on semantic consistency in an evolving vector field

Based on the Aristotelian concept of potentiality vs. actuality allowing for the study of energy and dynamics in language, we propose a field approach to lexical analysis. Falling back on the distributional hypothesis to statistically model word meaning, we used evolving fields as a metaphor to express time-dependent changes in a vector space model by a combination of random indexing and evolving self-organizing maps (ESOM). To monitor semantic drifts within the observation period, an experiment was carried out on the term space of a collection of 12.8 million Amazon book reviews. For evaluation, the semantic consistency of ESOM term clusters was compared with their respective neighbourhoods in WordNet, and contrasted with distances among term vectors by random indexing. We found that at 0.05 level of significance, the terms in the clusters showed a high level of semantic consistency. Tracking the drift of distributional patterns in the term space across time periods, we found that consistency decreased, but not at a statistically significant level. Our method is highly scalable, with interpretations in philosophy.

Digital Preservation in Grids and Clouds: A Middleware Approach

Digital preservation is the persistent archiving of digital assets for future access and reuse, irrespective of the underlying platform and software solutions. Existing preservation systems have a strong focus on Grids, but the advent of cloud technologies offers an attractive option. We describe a middleware system that enables a flexible choice between a Grid and a cloud for ad-hoc computations that arise during the execution of a preservation workflow and also for archiving digital objects. The choice between different infrastructures remains open during the lifecycle of the archive, ensuring a smooth switch between different solutions to accommodate the changing requirements of the organization that needs its digital assets preserved. We also offer insights on the costs, running times, and organizational issues of cloud computing, proving that the cloud alternative is particularly attractive for smaller organizations without access to a Grid or with limited IT infrastructure.