arXiv research

Paper proposes a SIMO DNN for indoor localization using Wi-Fi fingerprints.

One of the key technologies for future large-scale location-aware services covering a complex of multi-story buildings --- e.g., a big shopping mall and a university campus --- is a scalable indoor localization technique. In this paper, we report the current status of our investigation on the use of deep neural network…

The paper develops a theory for random forests, separating variance components and providing methods for estimating prediction intervals.

Trading floors need to be twice as deep as electronic markets to compete.

The relationships between braid ordering and the geometry of its closure is studied. We prove that if an essential closed surface $F$ in the complements of closed braid has relatively small genus with respect to the Dehornoy floor of the braid, $F$ is circular-foliated in a sense of Birman-Menasco's Braid foliation the…

Improved analysis for fair federated learning reduces dependence on noise floor.

We propose a novel methodology, forest floor, to visualize and interpret random forest (RF) models. RF is a popular and useful tool for non-linear multi-variate classification and regression, which yields a good trade-off between robustness (low variance) and adaptiveness (low bias). Direct interpretation of a RF model…

We study the portfolio selection problem of a long-run investor who is maximising the asymptotic growth rate of her expected utility. We show that, somewhat surprisingly, it is essentially not affected by introduction of a floor constraint which requires the wealth process to dominate a given benchmark at all times. We…

Consider an agent who enters a financial market on day t = 0 with an initial capital amount x. He invests this amount on stocks and the money market, and by day t = T, has generated a wealth W . He is given a convex class of probability measures (called scenarios) and a real-valued function (or floors) corresponding to…

We present novel empirical observations regarding how stochastic gradient descent (SGD) navigates the loss landscape of over-parametrized deep neural networks (DNNs). These observations expose the qualitatively different roles of learning rate and batch-size in DNN optimization and generalization. Specifically we study…

Study proposes explainable analytics for manufacturing process planning.

EdgeLite detects hazardous supermarket floors, improving safety.

Algorithm speeds up search for stationary targets with guaranteed accuracy.

We determine the price of digital double barrier options with an arbitrary number of barrier periods in the Black-Scholes model. This means that the barriers are active during some time intervals, but are switched off in between. As an application, we calculate the value of a structure floor for structured notes whose …

The study uses machine learning to analyze office floor plans and predict function based on geometry.

Enhances ocean floor mapping with adaptive uncertainty estimates.

Develops a new framework for perpetual futures on binary prediction markets.

A training-free conformal interval is a mandatory baseline for probabilistic time-series forecasting.

New network approximates functions with error decreasing with network width and depth.

Machine learning detects new minerals at Mars rover landing sites.

The study bounds exceptional surgeries for hyperbolic knots.

This paper considers the classification of linear subspaces with mismatched classifiers. In particular, we assume a model where one observes signals in the presence of isotropic Gaussian noise and the distribution of the signals conditioned on a given class is Gaussian with a zero mean and a low-rank covariance matrix.…

Trading strategies improved by classifying financial time-series images.

Three-hidden-layer neural networks can approximate Hölder continuous functions uniformly with exponential rate.

Language models allocate information storage, not collapsing into uniform representations.

Stochastic differential equation approximation for linear TD(0) under Markovian noise

Paper proposes RL for efficient dispatching in dynamic manufacturing environments.

New model estimates indoor radon distribution with higher spatial resolution.

Newly available data on the spatial distribution of retail activities in cities makes it possible to build models formalized at the level of the single retailer. Current models tackle consumer location choices at an aggregate level and the opportunity new data offers for modeling at the retail unit level lacks a theore…

The paper uses Bayesian Surprise to identify unexpected structures in indoor environments.

The genus of knots is a one of the fundamental invariant and can be seen as a complexity of knots. In this paper, we give a lower bound of genus using Dehornoy floor, which is a measure of complexity of braids in terms of braid ordering.

Deep RL policies can leak private information from trained policies.

AdaQuantFL reduces communication in federated learning by adaptively quantizing model updates.

A framework for cost of belief revision in uncertain agents.

Language models fail to process hallucinated responses, and this study diagnoses the failure.

The main result of this paper that a martingale evolution can be chosen for Libor such that all the Libor interest rates have a common market measure; the drift is fixed such that each Libor has the martingale property. Libor is described using a field theory model, and a common measure is seen to be emerge naturally f…

SAEs struggle with curved activation manifolds, revealing layer-dependent scaling laws.

We investigate the application of two heuristic methods, genetic algorithms and tabu/scatter search, to the optimisation of realistic portfolios. The model is based on the classical mean-variance approach, but enhanced with floor and ceiling constraints, cardinality constraints and nonlinear transaction costs which inc…

Fast, reliable, and error-bounded option pricing with neural networks

Optimal portfolio tracking with dynamic capital injection into a ratcheting benchmark.

Optimal solar energy production and trading strategies in SREC markets identified.

A study on the depth of graph neural networks on sparse graphs, revealing a dichotomy based on the Kesten-Stigum ratio.

Cumulant expansion is used to derive accurate closed-form approximation for Monthly Sum Options in case of constant volatility model. Payoff of Monthly Sum Option is based on sum of $N$ caped (and probably floored) returns. It is noticed, that $1/\sqrt{N}$ can be used as a small parameter in Edgeworth expansion. First …

The paper analyzes Adam and SGD in nonstationary optimization, revealing tradeoffs between noise and drift.

Unified framework for mean testing under truncation bias.

A new method selects models for ensemble learning to maximize mutual information, outperforming existing approaches.

The paper decomposes unsupervised learning's generalization error into model, data, and variance components.

TASID learns policies in high-dimensional settings with abstract simulator knowledge.