A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.
Paper tackles tensor decomposition for unaligned observations using RKHS and novel loss functions.
Training models to prefer certain responses can unintentionally shift probability to harmful ones.
A new method compares unaligned datasets using log-Euclidean signatures of SPD matrices.
We develop the multilingual topic model for unaligned text (MuTo), a probabilistic model of text that is designed to analyze corpora composed of documents in two languages. From these documents, MuTo uses stochastic EM to simultaneously discover both a matching between the languages and multilingual latent topics. We d…
We propose a new nonlinear factorization model for graphs that are with topological structures, and optionally, node attributes. This model is based on a pseudometric called Gromov-Wasserstein (GW) discrepancy, which compares graphs in a relational way. It estimates observed graphs as GW barycenters constructed by a se…
Despite the eminent successes of deep neural networks, many architectures are often hard to transfer to irregularly-sampled and asynchronous time series that commonly occur in real-world datasets, especially in healthcare applications. This paper proposes a novel approach for classifying irregularly-sampled time series…
Deep reinforcement learning methods traditionally struggle with tasks where environment rewards are particularly sparse. One successful method of guiding exploration in these domains is to imitate trajectories provided by a human demonstrator. However, these demonstrations are typically collected under artificial condi…
Extends reinforcement learning alignment to scalar rewards, improving math reasoning.
The recent increase in the scale and complexity of software systems has introduced new challenges to the time series monitoring and anomaly detection process. A major drawback of existing anomaly detection methods is that they lack contextual information to help stakeholders identify the cause of anomalies. This proble…
Aggregating multi-subject functional magnetic resonance imaging (fMRI) data is indispensable for generating valid and general inferences from patterns distributed across human brains. The disparities in anatomical structures and functional topographies of human brains warrant aligning fMRI data across subjects. However…
ATLAS separates invariant and transferable latent factors across diverse environments.
The ability to represent and compare machine learning models is crucial in order to quantify subtle model changes, evaluate generative models, and gather insights on neural network architectures. Existing techniques for comparing data distributions focus on global data properties such as mean and covariance; in that se…
Jukebox generates high-fidelity songs with singing in raw audio.
Clustering and community detection with multiple graphs have typically focused on aligned graphs, where there is a mapping between nodes across the graphs (e.g., multi-view, multi-layer, temporal graphs). However, there are numerous application areas with multiple graphs that are only partially aligned, or even unalign…
We propose a flexible framework for spectral conversion (SC) that facilitates training with unaligned corpora. Many SC frameworks require parallel corpora, phonetic alignments, or explicit frame-wise correspondence for learning conversion functions or for synthesizing a target spectrum with the aid of alignments. Howev…
LLM safety alignment explained as divergence estimation.
An increasing number of datasets contain multiple views, such as video, sound and automatic captions. A basic challenge in representation learning is how to leverage multiple views to learn better representations. This is further complicated by the existence of a latent alignment between views, such as between speech a…
TACTiS models time series uncertainty with transformer attention.
Proposes HOT method for robust multi-view learning.
Proposes a VAE with a discrete bottleneck for better text generation.
Improved model for multivariate time series prediction with simpler architecture.
Deep latent variable models (LVM) such as variational auto-encoder (VAE) have recently played an important role in text generation. One key factor is the exploitation of smooth latent structures to guide the generation. However, the representation power of VAEs is limited due to two reasons: (1) the Gaussian assumption…
Machine learning methods struggle with geometric data, but shape space analysis provides a framework for studying and analyzing geometric variability.
We study the question of how to imitate tasks across domains with discrepancies such as embodiment, viewpoint, and dynamics mismatch. Many prior works require paired, aligned demonstrations and an additional RL step that requires environment interactions. However, paired, aligned demonstrations are seldom obtainable an…
Proposes a model for multi-horizon probabilistic forecasting of time series influenced by asynchronous events.
Set-Sequence model learns cross-sectional dynamics directly from time series data.
A new method matches measures across different spaces using cost-regularized optimal transport.
New method identifies shared components from unpaired multimodal mixtures.
In practice, there are often explicit constraints on what representations or decisions are acceptable in an application of machine learning. For example it may be a legal requirement that a decision must not favour a particular group. Alternatively it can be that that representation of data must not have identifying in…
Unified model improves multi-task learning by accounting for temporal misalignment.
A new method constrains PARAFAC2 for better pattern recovery.
Gene annotation has traditionally required direct comparison of DNA sequences between an unknown gene and a database of known ones using string comparison methods. However, these methods do not provide useful information when a gene does not have a close match in the database. In addition, each comparison can be costly…
Paper identifies unobserved variables from observable data.
New framework learns policies for partially observable systems.
Efficient RL in partially observable risk-sensitive environments with hindsight observations.
We consider the problem of diagnosis where a set of simple observations are used to infer a potentially complex hidden hypothesis. Finding the optimal subset of observations is intractable in general, thus we focus on the problem of active diagnosis, where the agent selects the next most-informative observation based o…
The `observer space' of a Lorentzian spacetime is the space of future-timelike unit tangent vectors. Using Cartan geometry, we first study the structure a given spacetime induces on its observer space, then use this to define abstract observer space geometries for which no underlying spacetime is assumed. We propose ta…
A method uses CG to create efficient channels for ideal observers.
Extends PD-NJ-ODE to noisy observations and dependent observation times.
Uncorrelated optical space observation association represents a classic needle in a haystack problem. The objective being to find small groups of observations that are likely of the same resident space objects (RSOs) from amongst the much larger population of all uncorrelated observations. These observations being pote…
The paper analyzes the stability of an observer error in a vibrating string system.
New algorithm for aggregate inference in HMMs with continuous observations.
The paper compares and optimizes estimators for treatment effects with observed confounders and mediators.
New algorithm improves reinforcement learning from partial observations.
CVRL tackles complex visual observations in reinforcement learning.
A new algorithm CAP learns optimal policies from observational data with confounding bias and missing observations.
We study the problem of learning influence functions under incomplete observations of node activations. Incomplete observations are a major concern as most (online and real-world) social networks are not fully observable. We establish both proper and improper PAC learnability of influence functions under randomly missi…
New surface observables yield 2-knot invariants in nonabelian theories.