On-device research index

arXiv research

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.

169,291 papers · 148 categories

Trend · papers per month

count % of corpus

6 results for “Full-Capacity”

Full-Capacity Unitary Recurrent Neural Networksstat.ML

Full-capacity uRNNs improve performance over restricted-capacity ones.

problem Vanishing and exploding gradient issues in recurrent neural networks.

method Optimized full-capacity unitary recurrence matrices over all unitary matrices.

result Significantly improved performance compared to LSTMs and restricted-capacity uRNNs.

Recurrent Neural Networks Unitary Matrices Gradient Issues Full-Capacity Optimization

2016-10-31abs ↗pdf ↗

Exploiting the Full Capacity of Deep Neural Networks while Avoiding Overfitting by Targeted Sparsity Regularizationcs.LG

New method targets sparsity to prevent overfitting in deep nets.

problem Overfitting in deep neural networks with small datasets.

method Targeted sparsity regularization to visualize and counteract overfitting.

result Significant increase in image classification performance without overfitting.

deep learning overfitting sparsity regularization activation

2020-02-21abs ↗pdf ↗

Trading under the Proof-of-Stake Protocol -- a Continuous-Time Control Approachecon.GN

Optimal trading strategy in Proof-of-Stake blockchain using continuous-time control.

problem Finding the optimal balance between stake utility and consumption utility in Proof-of-Stake blockchain.

method Continuous-time control approach, dynamic programming, Hamilton-Jacobi-Bellman (HJB) equations.

result Close-form solutions for linear and convex utility functions, optimal strategies identified.

Proof-of-Stake Blockchain Optimal Trading Continuous-Time Control Dynamic Programming

2022-07-26abs ↗pdf ↗

Stochastic Estimated Risk for Storage Capacityq-fin.GN

SER tool predicts storage capacity risk with probabilistic accuracy.

problem Managing data storage growth to avoid financial losses.

method Stochastic Estimated Risk (SER) using Brownian motion with drift model.

result Probabilistic approach is more accurate for non-linear storage patterns.

storage capacity stochastic estimated risk data growth financial losses probabilistic accuracy

2018-12-01abs ↗pdf ↗

Parseval Networks: Improving Robustness to Adversarial Examplesstat.ML

Parseval networks improve deep nets' robustness to adversarial examples.

problem Improving deep neural networks' robustness to adversarial attacks.

method Constraining the Lipschitz constant and maintaining Parseval tight frames in weight matrices.

result Parseval networks maintain accuracy and robustness to adversarial examples compared to vanilla networks.

adversarial robustness Parseval tight frames Lipschitz constant deep learning convolutional neural networks

2017-04-28abs ↗pdf ↗

India Growth Forecast for 2020-21q-fin.GN

India's 2020-21 GDP growth forecast is projected at 1.9% due to COVID-19.

problem Impact of COVID-19 on India's 2020-21 GDP growth forecast.

method Quarterly GVA estimates using capacity utilisation assumptions and 2019-20 data.

result India's GDP growth in 2020-21 is expected to be 1.9%.

India GDP GVA COVID-19 Economic Forecast Capacity Utilisation

2020-05-14abs ↗pdf ↗