A parallel Fortran framework for neural networks and deep learning.
problem Developing efficient parallel Fortran for neural networks and deep learning.
method Simple interface, activation functions, stochastic gradient descent, Fortran 2018 collective subroutines, parallelism with derived types and collective operations.
result Ease of use and computational performance similar to existing machine learning frameworks, suitable for production.
The realized stochastic volatility (RSV) model that utilizes the realized volatility as additional information has been proposed to infer volatility of financial time series. We consider the Bayesian inference of the RSV model by the Hybrid Monte Carlo (HMC) algorithm. The HMC algorithm can be parallelized and thus per…
ParaMonte simplifies Monte Carlo simulations for various scientific fields.
problem Efficiently performing Monte Carlo simulations for complex models.
method Unified, high-performance, parallelized library for C, C++, Fortran.
result Automates and streamlines Monte Carlo sampling for arbitrary-dimensional functions.
Improved model error correction online with neural networks in 4D-Var.
problem Reconstructing dynamics of imperfectly observed physical models.
method Weak-constraint 4D-Var framework with online neural network training.
result Online model error correction yields more accurate results than offline.
ICML workshop on making machine learning models more understandable.
problem Making machine learning models more understandable to humans.
method Various presentations and discussions on interpretability techniques.
result Advancements in methods to improve model interpretability.
ML4H workshop at NeurIPS 2018 focuses on health applications of machine learning.
problem Improving health outcomes through machine learning.
method Presented health applications and machine learning techniques.
result Demonstrated the potential of machine learning in healthcare.
NIPS 2018 Adversarial Vision Challenge aims to improve machine vision models.
problem Improving robustness of machine vision models and adversarial attacks.
method Organized a competition to measure progress in adversarial machine vision.
result Facilitated measurable progress in adversarial machine vision.
Workshop on ML for developing countries to enhance sustainability.
problem Addressing challenges in machine learning for developing nations.
method Discussion and collaboration among researchers.
result Promoting sustainable impact through machine learning.
Mask-RCNN applied to ISIC 2018 lesion tasks.
problem Lesion boundary segmentation, attributes detection, and diagnosis.
method Mask-RCNN applied to ISIC 2018 challenge tasks with a trained model for task 1 and a simple voting procedure for task 3.
result Improved lesion diagnosis accuracy using Mask-RCNN.
SimTensor is a multi-platform, open-source software for generating artificial tensor data (either with CP/PARAFAC or Tucker structure) for reproducible research on tensor factorization algorithms. SimTensor is a stand-alone application based on MATALB. It provides a wide range of facilities for generating tensor data w…
Paper improves ATN for generating adversarial examples.
problem Generating adversarial examples to fool models.
method Improves Adversarial Transformation Networks (ATN).
result Won 2nd place in CAAD 2018 non-target task.
New method improves fairness in machine learning models.
problem Reducing bias in machine learning classification models.
method Formulated as multi-objective optimization, uses gradient descent-ascent algorithm with modified gradient update step.
result Empirical tests show improved fairness compared to state-of-the-art algorithms.
One advantage of decision tree based methods like random forests is their ability to natively handle categorical predictors without having to first transform them (e.g., by using feature engineering techniques). However, in this paper, we show how this capability can lead to an inherent "absent levels" problem for deci…
Neural networks are known to be vulnerable to adversarial examples. In this note, we evaluate the two white-box defenses that appeared at CVPR 2018 and find they are ineffective: when applying existing techniques, we can reduce the accuracy of the defended models to 0%.
ARIMA and LSTM models predict stock prices with varying accuracy.
problem Predicting stock prices accurately.
method ARIMA and LSTM models trained on historical stock data.
result LSTM model performs better in predicting stock price changes.
Amobee wins WASSA 2018 emotion prediction with language models and LSTM.
problem Predicting emotions from tweets without explicit mentions.
method Ensemble system of language models and LSTM with CNN attention.
result 1st place with macro F1 score of 0.7145.
FOLKLORE algorithm speeds up online multiclass logistic regression.
problem Efficiently solving online multiclass logistic regression without high computational cost.
method Developed FOLKLORE algorithm with improved runtime and regret bound.
result First practical algorithm for online multiclass logistic regression.
Improved performance in classifying domestic activities.
problem Classifying domestic activities effectively.
method Ensemble learning system based on CNN and LSTM.
result Significant improvement in F1-score (92.19% vs baseline 84.49%).
MedAusbild team won ISIC challenge by classifying seven skin diseases.
problem Skin disease classification using medical images.
method Employed deep learning, specifically convolutional neural networks (CNNs), for seven skin disease classification.
result MedAusbild team won ISIC challenge 2018.
Bayesian approach models policy distribution for faster exploration and transfer learning.
problem Challenges in exploration and adaptation to new tasks in transfer learning.
method Modeling a distribution over policies in a Bayesian deep reinforcement learning setup.
result Favorable experimental results on GridWorld and MiniGrid environments.
Deep reinforcement learning improves forex trading by handling complex, random processes.
problem Stable trends in deep learning predictions for forex trading.
method Used reinforcement learning, optimized Sure-Fire policy, encoded price data, compared DQN and PPO.
result Models achieved favorable investment performance, validating reinforcement learning feasibility.
Paper derives a fast learning rate for deep neural networks without scale invariant activation functions.
problem Analyzing the impact of non-scale invariant activation functions on deep learning performance.
method Using Suzuki (2018) framework, derived a tight generalization error bound for deep neural networks with non-scale invariant activations.
result Without scale invariance of activation functions, deep learning can still achieve a fast learning rate.
Improved path-length regret bounds for adaptive and oblivious adversaries.
problem Adaptive and oblivious adversaries in multi-armed bandit and linear bandit problems.
method Developed two new algorithms based on optimistic mirror descent framework with novel techniques.
result Strictly improved path-length bounds for adaptive adversary and better results for oblivious adversary.
New interpolation methods outperform Gaussian smoothing in derivative-free optimization.
problem Optimizing functions with noisy evaluations and no derivative information.
method First-order line search methods using linear interpolation vs. Gaussian smoothing.
result Linear interpolation yields better convergence rates and performance.
Study analyzes Airbnb booking lead times during global crises using a new metric.
problem Disruptions in booking behaviors during global crises affect forecasting accuracy.
method Normalized L1 (Manhattan) distance to assess lead time divergences.
result Identified two-phase disruption: abrupt change at pandemic onset followed by partial recovery.
Deep learning model classifies gastrointestinal diseases with high accuracy.
problem Disease detection in the gastrointestinal tract.
method Global features and deep neural networks.
result 95.80% accuracy, 95.87% precision, 95.80% F1-score.
This paper describes the participation of Amobee in the shared sentiment analysis task at SemEval 2018. We participated in all the English sub-tasks and the Spanish valence tasks. Our system consists of three parts: training task-specific word embeddings, training a model consisting of gated-recurrent-units (GRU) with …
New flaw found in SAP defense, reducing its effectiveness to 0.1%.
problem Weakness in Stochastic Activation Pruning defense against adversarial attacks.
method Re-examined the implementation of SAP and introduced a new BPDA attack.
result SAP's effectiveness reduced to 0.1% when properly applied.
We consider a team of reinforcement learning agents that concurrently operate in a common environment, and we develop an approach to efficient coordinated exploration that is suitable for problems of practical scale. Our approach builds on seed sampling (Dimakopoulou and Van Roy, 2018) and randomized value function lea…
Task describes tagging audio clips from 41 categories.
problem Tagging audio clips from diverse categories.
method Baseline system using AudioSet labels.
result Demonstrates feasibility of general-purpose audio tagging.
AutoML challenge solved lifelong learning problems without i.i.d. data.
problem Designing lifelong learning systems without independent and identically distributed data.
method Developed and evaluated machine learning programs using CodaLab platform.
result More than 300 participants competed and solved complex lifelong learning problems.
Cheap model diagnoses vocal disorders accurately.
problem Diagnosing vocal disorders without expensive equipment.
method Used Mel-Cepstrum vectors and Support Vector Machine.
result Accurately diagnosed three vocal disorders.
In this paper, we develop improved techniques for defending against adversarial examples at scale. First, we implement the state of the art version of adversarial training at unprecedented scale on ImageNet and investigate whether it remains effective in this setting - an important open scientific question (Athalye et …
Study uses satellite data to predict tailings dam collapse risk.
problem Detecting early signs of tailings dam instability.
method Spectral analysis of satellite InSAR displacement time series data.
result Algorithm detects risk milestones up to 5 months before dam collapse.
ParaDRAM automates parallel MCMC simulations across languages.
problem Efficiently sampling complex mathematical functions.
method Automated, parallel, high-performance MCMC with adaptive rejection.
result Significant memory reduction for complex simulations.
We show that finite-width deep ReLU neural networks yield rate-distortion optimal approximation (Bölcskei et al., 2018) of polynomials, windowed sinusoidal functions, one-dimensional oscillatory textures, and the Weierstrass function, a fractal function which is continuous but nowhere differentiable. Together with thei…
Simpler proof for Kielak's virtual fibering criterion.
problem Virtual fibering criterion for RFRS groups
method Simpler proof
result Simplified proof of Kielak's criterion
This paper establishes lower bounds for smooth nonconvex finite-sum optimization.
problem Understanding the complexity of finding optimal solutions in nonconvex finite-sum optimization.
method Proving tight lower bounds for the complexity of finding ε-suboptimal points and ε-approximate stationary points.
result Existing algorithms achieve optimal IFO complexity up to logarithmic factors.
Classifies isotopy classes of links from Thompson's group F and its subgroup.
problem Classify isotopy classes of links from Thompson's group F and its subgroup.
method Introduced a method to produce links from elements of Thompson's group F and its subgroup.
result Classified isotopy classes of links from Thompson's group F and its subgroup.
This Chapter, "High-dimensional ABC", is to appear in the forthcoming Handbook of Approximate Bayesian Computation (2018). It details the main ideas and concepts behind extending ABC methods to higher dimensions, with supporting examples and illustrations.
We re-examine and extend the findings from the recent paper by Dumitrescu, Quenez and Sulem (2018) who studied American and game options in a particular market model using the nonlinear arbitrage-free pricing approach developed in El Karoui and Quenez (1997). In the first part, we provide a detailed study of unilateral…
Proposes a stochastic model for South African actuarial use.
problem Long-term forecasting for South African institutions.
method Modeling economic series, estimating parameters, testing stability.
result Validated model for long-term forecasts.
Improved BERT model with latent persona and topic variables.
problem Improving BERT's domain-specific utility while maintaining generalization.
method Combining BERT with Universal Transformer, adding latent persona and topic variables.
result Pre-trained model for social texts outperforms baseline.
ES and FD gradients converge as optimization dimension grows.
problem Understanding the relationship between Evolution Strategies and Finite Differences gradients.
method Analyzing the convergence of gradients as the optimization dimension increases.
result ES and FD gradients converge as the dimension of the vector under optimization increases.
On the fifth of February, 2018, the Dow Jones Industrial Average dropped 1,175.21 points, the largest single-day fall in history in raw point terms. This followed a 666-point loss on the second, and another drop of over a thousand points occurred three days later. It is natural to ask whether these events indicate a tr…
This Chapter, "Overview of Approximate Bayesian Computation", is to appear as the first chapter in the forthcoming Handbook of Approximate Bayesian Computation (2018). It details the main ideas and concepts behind ABC methods with many examples and illustrations.
The study uses LSTM and random forests to forecast stock price movements for intraday trading.
problem Forecasting directional movements of stock prices for intraday trading.
method Employed random forests and LSTM networks to analyze S&P 500 constituent stocks.
result Multi-feature setting provided higher daily returns (0.64% using LSTM, 0.54% using random forests) compared to single-feature setting.
Unsupervised machine translation---i.e., not assuming any cross-lingual supervision signal, whether a dictionary, translations, or comparable corpora---seems impossible, but nevertheless, Lample et al. (2018) recently proposed a fully unsupervised machine translation (MT) model. The model relies heavily on an adversari…