arXiv subjects

Jonas Fischer

Publications and source records attributed to Jonas Fischer.

31 records · Page 2Linked to original sources

arXiv2024

Sailing in high-dimensional spaces: Low-dimensional embeddings through angle preservation

Low-dimensional embeddings (LDEs) of high-dimensional data are ubiquitous in science and engineering. They allow us to quickly understand the main properties of the data, identify outliers and processing errors, and inform the next steps of data analysis. As such, LDEs have to be faithful to the original high-dimensional data, i.e., they should represent the relationships that are encoded in the data, both at a local as well as global scale. The current generation of LDE approaches focus on reconstructing local distances between any pair of samples correctly, often out-performing traditional approaches aiming at all distances. For these approaches, global relationships are, however, usually strongly distorted, often argued to be an inherent trade-off between local and global structure learning for embeddings. We suggest a new perspective on LDE learning, reconstructing angles between data points. We show that this approach, Mercat, yields good reconstruction across a diverse set of experiments and metrics, and preserve structures well across all scales. Compared to existing work, our approach also has a simple formulation, facilitating future theoretical analysis and algorithmic improvements.

cs.LG↗

arXiv2023

Finding Interpretable Class-Specific Patterns through Efficient Neural Search

Discovering patterns in data that best describe the differences between classes allows to hypothesize and reason about class-specific mechanisms. In molecular biology, for example, this bears promise of advancing the understanding of cellular processes differing between tissues or diseases, which could lead to novel treatments. To be useful in practice, methods that tackle the problem of finding such differential patterns have to be readily interpretable by domain experts, and scalable to the extremely high-dimensional data. In this work, we propose a novel, inherently interpretable binary neural network architecture DIFFNAPS that extracts differential patterns from data. DiffNaps is scalable to hundreds of thousands of features and robust to noise, thus overcoming the limitations of current state-of-the-art methods in large-scale applications such as in biology. We show on synthetic and real world data, including three biological applications, that, unlike its competitors, DiffNaps consistently yields accurate, succinct, and interpretable class descriptions

cs.LG↗

arXiv2023

Understanding and Mitigating Classification Errors Through Interpretable Token Patterns

State-of-the-art NLP methods achieve human-like performance on many tasks, but make errors nevertheless. Characterizing these errors in easily interpretable terms gives insight into whether a classifier is prone to making systematic errors, but also gives a way to act and improve the classifier. We propose to discover those patterns of tokens that distinguish correct and erroneous predictions as to obtain global and interpretable descriptions for arbitrary NLP classifiers. We formulate the problem of finding a succinct and non-redundant set of such patterns in terms of the Minimum Description Length principle. Through an extensive set of experiments, we show that our method, Premise, performs well in practice. Unlike existing solutions, it recovers ground truth, even on highly imbalanced data over large vocabularies. In VQA and NER case studies, we confirm that it gives clear and actionable insight into the systematic errors made by NLP classifiers.

cs.CL↗

arXiv2023

Federated Learning from Small Datasets

Federated learning allows multiple parties to collaboratively train a joint model without sharing local data. This enables applications of machine learning in settings of inherently distributed, undisclosable data such as in the medical domain. In practice, joint training is usually achieved by aggregating local models, for which local training objectives have to be in expectation similar to the joint (global) objective. Often, however, local datasets are so small that local objectives differ greatly from the global objective, resulting in federated learning to fail. We propose a novel approach that intertwines model aggregations with permutations of local models. The permutations expose each local model to a daisy chain of local datasets resulting in more efficient training in data-sparse domains. This enables training on extremely small local datasets, such as patient data across hospitals, while retaining the training efficiency and privacy benefits of federated learning.

cs.LG↗

arXiv2023

Preserving local densities in low-dimensional embeddings

Low-dimensional embeddings and visualizations are an indispensable tool for analysis of high-dimensional data. State-of-the-art methods, such as tSNE and UMAP, excel in unveiling local structures hidden in high-dimensional data and are therefore routinely applied in standard analysis pipelines in biology. We show, however, that these methods fail to reconstruct local properties, such as relative differences in densities (Fig. 1) and that apparent differences in cluster size can arise from computational artifact caused by differing sample sizes (Fig. 2). Providing a theoretical analysis of this issue, we then suggest dtSNE, which approximately conserves local densities. In an extensive study on synthetic benchmark and real world data comparing against five state-of-the-art methods, we empirically show that dtSNE provides similar global reconstruction, but yields much more accurate depictions of local distances and relative densities.

cs.LG↗

arXiv2022

Optimal control and selectivity of qubits in contact with a structured environment

We present a theoretical study of the optimal control of a qubit interacting with a structured environment. We consider a model system in which the bath is a bosonic reservoir at zero temperature and the qubit frequency is the only control parameter. Using optimal control techniques, we show the extent to which qubit population and relaxation effects can be manipulated. The reachable qubit states by a shaped control with a fixed maximum intensity are found numerically. We analyze the role of standard control mechanisms and the structure of the set of reachable states with respect to the coupling strength between the system and the environment. This investigation is used as a starting point to explore the selectivity problem of two uncoupled qubits interacting with their own baths and characterized by a specific coupling strength. We numerically derive the optimal control solution for a wide range of parameters and we show that the control law is close to a sinusoidal function with a specific frequency in some peculiar cases.

quant-ph↗

arXiv2022

Label-Descriptive Patterns and Their Application to Characterizing Classification Errors

State-of-the-art deep learning methods achieve human-like performance on many tasks, but make errors nevertheless. Characterizing these errors in easily interpretable terms gives insight into whether a classifier is prone to making systematic errors, but also gives a way to act and improve the classifier. We propose to discover those feature-value combinations (i.e., patterns) that strongly correlate with correct resp. erroneous predictions to obtain a global and interpretable description for arbitrary classifiers. We show this is an instance of the more general label description problem, which we formulate in terms of the Minimum Description Length principle. To discover a good pattern set, we develop the efficient Premise algorithm. Through an extensive set of experiments we show it performs very well in practice on both synthetic and real-world data. Unlike existing solutions, it ably recovers ground truth patterns, even on highly imbalanced data over many features. Through two case studies on Visual Question Answering and Named Entity Recognition, we confirm that Premise gives clear and actionable insight into the systematic errors made by modern NLP classifiers.

cs.LG↗

arXiv2022

Lottery Tickets with Nonzero Biases

The strong lottery ticket hypothesis holds the promise that pruning randomly initialized deep neural networks could offer a computationally efficient alternative to deep learning with stochastic gradient descent. Common parameter initialization schemes and existence proofs, however, are focused on networks with zero biases, thus foregoing the potential universal approximation property of pruning. To fill this gap, we extend multiple initialization schemes and existence proofs to nonzero biases, including explicit 'looks-linear' approaches for ReLU activation functions. These do not only enable truly orthogonal parameter initialization but also reduce potential pruning errors. In experiments on standard benchmark data, we further highlight the practical benefits of nonzero bias initialization schemes, and present theoretically inspired extensions for state-of-the-art strong lottery ticket pruning.

cs.LG↗

arXiv2022

Plant 'n' Seek: Can You Find the Winning Ticket?

The lottery ticket hypothesis has sparked the rapid development of pruning algorithms that aim to reduce the computational costs associated with deep learning during training and model deployment. Currently, such algorithms are primarily evaluated on imaging data, for which we lack ground truth information and thus the understanding of how sparse lottery tickets could be. To fill this gap, we develop a framework that allows us to plant and hide winning tickets with desirable properties in randomly initialized neural networks. To analyze the ability of state-of-the-art pruning to identify tickets of extreme sparsity, we design and hide such tickets solving four challenging tasks. In extensive experiments, we observe similar trends as in imaging studies, indicating that our framework can provide transferable insights into realistic problems. Additionally, we can now see beyond such relative trends and highlight limitations of current pruning methods. Based on our results, we conclude that the current limitations in ticket sparsity are likely of algorithmic rather than fundamental nature. We anticipate that comparisons to planted tickets will facilitate future developments of efficient pruning algorithms.

cs.LG↗

arXiv2022

Estimating Mutual Information via Geodesic $k$NN

Estimating mutual information (MI) between two continuous random variables $X$ and $Y$ allows to capture non-linear dependencies between them, non-parametrically. As such, MI estimation lies at the core of many data science applications. Yet, robustly estimating MI for high-dimensional $X$ and $Y$ is still an open research question. In this paper, we formulate this problem through the lens of manifold learning. That is, we leverage the common assumption that the information of $X$ and $Y$ is captured by a low-dimensional manifold embedded in the observed high-dimensional space and transfer it to MI estimation. As an extension to state-of-the-art $k$NN estimators, we propose to determine the $k$-nearest neighbors via geodesic distances on this manifold rather than from the ambient space, which allows us to estimate MI even in the high-dimensional setting. An empirical evaluation of our method, G-KSG, against the state-of-the-art shows that it yields good estimations of MI in classical benchmark and manifold tasks, even for high dimensional datasets, which none of the existing methods can provide.

cs.IT↗

arXiv2021

Factoring out prior knowledge from low-dimensional embeddings

Low-dimensional embedding techniques such as tSNE and UMAP allow visualizing high-dimensional data and therewith facilitate the discovery of interesting structure. Although they are widely used, they visualize data as is, rather than in light of the background knowledge we have about the data. What we already know, however, strongly determines what is novel and hence interesting. In this paper we propose two methods for factoring out prior knowledge in the form of distance matrices from low-dimensional embeddings. To factor out prior knowledge from tSNE embeddings, we propose JEDI that adapts the tSNE objective in a principled way using Jensen-Shannon divergence. To factor out prior knowledge from any downstream embedding approach, we propose CONFETTI, in which we directly operate on the input distance matrices. Extensive experiments on both synthetic and real world data show that both methods work well, providing embeddings that exhibit meaningful structure that would otherwise remain hidden.

cs.LG↗

arXiv2020

Fundamental Bounds on Qubit Reset

Qubit reset is a basic prerequisite for operating quantum devices, requiring the export of entropy. The fastest and most accurate way to reset a qubit is obtained by coupling the qubit to an ancilla on demand. Here, we derive fundamental bounds on qubit reset in terms of maximum fidelity and minimum time, assuming control over the qubit and no control over the ancilla. Using the Cartan decomposition of the Lie algebra of qubit plus two-level ancilla, we identify the types of interaction and controls for which the qubit can be purified. For these configurations, we show that a time-optimal protocol consists of purity exchange between qubit and ancilla brought into resonance, where the maximum fidelity is identical for all cases but the minimum time depends on the type of interaction and control. Furthermore, we find the maximally achievable fidelity to increase with the size of the ancilla Hilbert space, whereas the reset time remains constant.

quant-ph↗

arXiv2019

Time-optimal purification of a qubit in contact with a structured environment

We investigate the time-optimal control of the purification of a qubit interacting with a structured environment, consisting of a strongly coupled two-level defect in interaction with a thermal bath. On the basis of a geometric analysis, we show for weak and strong interaction strengths that the optimal control strategy corresponds to a qubit in resonance with the reservoir mode. We investigate when qubit coherence and correlation between the qubit and the environment speed-up the control process.

quant-ph↗