Data-driven interatomic potentials have emerged as a powerful class of surrogate models for {\it ab initio} potential energy surfaces that are able to reliably predict macroscopic properties with experimental accuracy. In generating accurate and transferable potentials the most time-consuming and arguably most important task is generating the training set, which still requires significant expert user input. To accelerate this process, this work presents \text{\it hyperactive learning} (HAL), a framework for formulating an accelerated sampling algorithm specifically for the task of training database generation. The key idea is to start from a physically motivated sampler (e.g., molecular dynamics) and add a biasing term that drives the system towards high uncertainty and thus to unseen training configurations. Building on this framework, general protocols for building training databases for alloys and polymers leveraging the HAL framework will be presented. For alloys, ACE potentials for AlSi10 are created by fitting to a minimal HAL-generated database containing 88 configurations (32 atoms each) with fast evaluation times of <100 microsecond/atom/cpu-core. These potentials are demonstrated to predict the melting temperature with excellent accuracy. For polymers, a HAL database is built using ACE, able to determine the density of a long polyethylene glycol (PEG) polymer formed of 200 monomer units with experimental accuracy by only fitting to small isolated PEG polymers with sizes ranging from 2 to 32.

Unsupervised representation learning aims at describing raw data efficiently to solve various downstream tasks. It has been approached with many techniques, such as manifold learning, diffusion maps, or more recently self-supervised learning. Those techniques are arguably all based on the underlying assumption that target functions, associated with future downstream tasks, have low variations in densely populated regions of the input space. Unveiling minimal variations as a guiding principle behind unsupervised representation learning paves the way to better practical guidelines for self-supervised learning algorithms.

Chain event graphs (CEGs) are a family of probabilistic graphical models that were first proposed in Smith and Anderson (2008) as an alternative to the family of Bayesian networks (BNs). In particular, CEGs were developed to explicitly accommodate processes exhibiting asymmetries of two types: (1) asymmetric independence structures or context-specific conditional independences where some statistical independences hold for certain values of the conditioning variables but not the others; and (2) asymmetric event spaces which are precisely event spaces that do not admit a product space structure. The latter asymmetry arises due to the presence of structural zeros and structural missing values, often-times by design (Shenvi & Smith, 2020). For example, consider modelling hospitalisations arising from infection caused by a circulating virus, and suppose that one of the two strains (call it strain A) of the virus has no treatment currently available while the other has a choice of two possible treatments. On the one hand, a variable of "Treatment" with state space {Treatment 1, Treatment 2} would be structurally missing and have no sensible value for those infected by strain A of the virus. Whereas on the other hand, if its state space is redefined to be {Treatment 1, Treatment 2, No treatment} then Treatment 1 and Treatment 2 would have structurally zero counts for those infected by strain A, i.e. irrespective of the sample size, there would always be zero individuals who are treated with either Treatment 1 or Treatment 2 among those infected by strain A. Such a process is inherently asymmetric. BNs, being variable-based - i.e. they use variables as the building blocks of their models - are unable to fully describe such asymmetries within their underlying statistical model and graphical structure.

Fast transforms correspond to factorizations of the form $\mathbf{Z} = \mathbf{X}^{(1)} \ldots \mathbf{X}^{(J)}$, where each factor $ \mathbf{X}^{(\ell)}$ is sparse and possibly structured. This paper investigates essential uniqueness of such factorizations, i.e., uniqueness up to unavoidable scaling ambiguities. Our main contribution is to prove that any $N \times N$ matrix having the so-called butterfly structure admits an essentially unique factorization into $J$ butterfly factors (where $N = 2^{J}$), and that the factors can be recovered by a hierarchical factorization method, which consists in recursively factorizing the considered matrix into two factors. This hierarchical identifiability property relies on a simple identifiability condition in the two-layer and fixed-support setting. This approach contrasts with existing ones that fit the product of butterfly factors to a given matrix via gradient descent. The proposed method can be applied in particular to retrieve the factorization of the Hadamard or the discrete Fourier transform matrices of size $N=2^J$. Computing such factorizations costs $\mathcal{O}(N^{2})$, which is of the order of dense matrix-vector multiplication, while the obtained factorizations enable fast $\mathcal{O}(N \log N)$ matrix-vector multiplications and have the potential to be applied to compress deep neural networks.

Zebrafish are a common model organism used to identify new disease therapeutics. High-throughput drug screens can be performed on larval zebrafish in multi-well plates by observing changes in behaviour following a treatment. Analysis of this behaviour can be difficult, however, due to the high dimensionality of the data obtained. Statistical analysis of individual statistics (such as the distance travelled) is generally not powerful enough to detect meaningful differences between treatment groups. Here, we propose a method for classifying zebrafish models of Parkinson's disease by genotype at 5 days old. Using a set of 2D behavioural features, we train a multi-layer perceptron neural network. We further show that the use of integrated gradients can give insight into the impact of each behaviour feature on genotype classifications by the model. In this way, we provide a novel pipeline for classifying zebrafish larvae, beginning with feature preparation and ending with an impact analysis of said features.

In continual learning (CL), the goal is to design models that can learn a sequence of tasks without catastrophic forgetting. While there is a rich set of techniques for CL, relatively little understanding exists on how representations built by previous tasks benefit new tasks that are added to the network. To address this, we study the problem of continual representation learning (CRL) where we learn an evolving representation as new tasks arrive. Focusing on zero-forgetting methods where tasks are embedded in subnetworks (e.g., PackNet), we first provide experiments demonstrating CRL can significantly boost sample efficiency when learning new tasks. To explain this, we establish theoretical guarantees for CRL by providing sample complexity and generalization error bounds for new tasks by formalizing the statistical benefits of previously-learned representations. Our analysis and experiments also highlight the importance of the order in which we learn the tasks. Specifically, we show that CL benefits if the initial tasks have large sample size and high "representation diversity". Diversity ensures that adding new tasks incurs small representation mismatch and can be learned with few samples while training only few additional nonzero weights. Finally, we ask whether one can ensure each task subnetwork to be efficient during inference time while retaining the benefits of representation learning. To this end, we propose an inference-efficient variation of PackNet called Efficient Sparse PackNet (ESPN) which employs joint channel & weight pruning. ESPN embeds tasks in channel-sparse subnets requiring up to 80% less FLOPs to compute while approximately retaining accuracy and is very competitive with a variety of baselines. In summary, this work takes a step towards data and compute-efficient CL with a representation learning perspective. GitHub page: https://github.com/ucr-optml/CtRL

We introduce pyGSL, a Python library that provides efficient implementations of state-of-the-art graph structure learning models along with diverse datasets to evaluate them on. The implementations are written in GPU-friendly ways, allowing one to scale to much larger network tasks. A common interface is introduced for algorithm unrolling methods, unifying implementations of recent state-of-the-art techniques and allowing new methods to be quickly developed by avoiding the need to rebuild the underlying unrolling infrastructure. Implementations of differentiable graph structure learning models are written in PyTorch, allowing us to leverage the rich software ecosystem that exists e.g., around logging, hyperparameter search, and GPU-communication. This also makes it easy to incorporate these models as components in larger gradient based learning systems where differentiable estimates of graph structure may be useful, e.g. in latent graph learning. Diverse datasets and performance metrics allow consistent comparisons across models in this fast growing field. The full code repository can be found on https://github.com/maxwass/pyGSL.

We propose Okapi, a simple, efficient, and general method for robust semi-supervised learning based on online statistical matching. Our method uses a nearest-neighbours-based matching procedure to generate cross-domain views for a consistency loss, while eliminating statistical outliers. In order to perform the online matching in a runtime- and memory-efficient way, we draw upon the self-supervised literature and combine a memory bank with a slow-moving momentum encoder. The consistency loss is applied within the feature space, rather than on the predictive distribution, making the method agnostic to both the modality and the task in question. We experiment on the WILDS 2.0 datasets Sagawa et al., which significantly expands the range of modalities, applications, and shifts available for studying and benchmarking real-world unsupervised adaptation. Contrary to Sagawa et al., we show that it is in fact possible to leverage additional unlabelled data to improve upon empirical risk minimisation (ERM) results with the right method. Our method outperforms the baseline methods in terms of out-of-distribution (OOD) generalisation on the iWildCam (a multi-class classification task) and PovertyMap (a regression task) image datasets as well as the CivilComments (a binary classification task) text dataset. Furthermore, from a qualitative perspective, we show the matches obtained from the learned encoder are strongly semantically related. Code for our paper is publicly available at https://github.com/wearepal/okapi/.

Word translation without parallel corpora has become feasible, rivaling the performance of supervised methods. Recent findings have shown that the accuracy and robustness of unsupervised word translation (UWT) can be improved by making use of visual observations, which are universal representations across languages. In this work, we investigate the potential of using not only visual observations but also pretrained language-image models for enabling a more efficient and robust UWT. Specifically, we develop a novel UWT method dubbed Word Alignment using Language-Image Pretraining (WALIP), which leverages visual observations via the shared embedding space of images and texts provided by CLIP models (Radford et al., 2021). WALIP has a two-step procedure. First, we retrieve word pairs with high confidences of similarity, computed using our proposed image-based fingerprints, which define the initial pivot for the word alignment. Second, we apply our robust Procrustes algorithm to estimate the linear mapping between two embedding spaces, which iteratively corrects and refines the estimated alignment. Our extensive experiments show that WALIP improves upon the state-of-the-art performance of bilingual word alignment for a few language pairs across different word embeddings and displays great robustness to the dissimilarity of language pairs or training corpora for two word embeddings.

Transfer Learning is an area of statistics and machine learning research that seeks answers to the following question: how do we build successful learning algorithms when the data available for training our model is qualitatively different from the data we hope the model will perform well on? In this thesis, we focus on a specific area of Transfer Learning called label shift, also known as quantification. In quantification, the aforementioned discrepancy is isolated to a shift in the distribution of the response variable. In such a setting, accurately inferring the response variable's new distribution is both an important estimation task in its own right and a crucial step for ensuring that the learning algorithm can adapt to the new data. We make two contributions to this field. First, we present a new procedure called SELSE which estimates the shift in the response variable's distribution. Second, we prove that SELSE is semiparametric efficient among a large family of quantification algorithms, i.e., SELSE's normalized error has the smallest possible asymptotic variance matrix compared to any other algorithm in that family. This family includes nearly all existing algorithms, including ACC/PACC quantifiers and maximum likelihood based quantifiers such as EMQ and MLLS. Empirical experiments reveal that SELSE is competitive with, and in many cases outperforms, existing state-of-the-art quantification methods, and that this improvement is especially large when the number of test samples is far greater than the number of train samples.