iPrescribe offers fast online offer recommendations using deep learning.
problem Online offer recommendation in real-time.
method Ensemble of deep learning and machine learning algorithms, optimized streaming technology stack, and efficient LSTM deployment.
result 90th percentile recommendation latency of 38 milliseconds.
Bayesian framework optimizes 3D view selection for specific tasks.
problem Optimizing 3D view selection for specific tasks in reconstruction.
method Bayesian decision theory, prior and posterior distributions, stochastic surface reconstruction.
result Framework achieves superior performance with fewer views.
Next-gen reservoir computing models dynamical systems from time-series data.
problem Modeling dynamical systems from time-series data.
method Pseudorandom nonlinear projection of time-delay embedded inputs.
result Models remain stable over long rollouts and generalize beyond training data.
Proving that next-token prediction makes language models generate coherent long documents.
problem Understanding why language models generate coherent documents despite focusing on next-token prediction.
method Proving the power of next-token prediction in learning longer-range structure using Recurrent Neural Networks (RNN).
result Optimizing next-token prediction in RNNs yields a model that closely approximates the training distribution, even for long-range coherence.
Model predicts next destination for users based on past trips and features.
problem Predicting the next destination in multi-destination trips.
method Used Cleora for city graph embedding and EMDE for prediction.
result Achieved 2nd place in Booking Data Challenge.
The algorithmic trading comes from digitalisation of the processing of trading assets on financial markets. Since 1980 the computerization of the stock market offers real time processing of financial information. This technological revolution has offered processes and mathematic methods to identify best return on trans…
Law explains how LLMs learn to predict next tokens.
problem Understanding how LLMs process input data internally.
method Introduced a precise law governing token embeddings in LLMs.
result Each layer equally contributes to next-token prediction accuracy.
Autoregressive generative models consistently achieve the best results in density estimation tasks involving high dimensional data, such as images or audio. They pose density estimation as a sequence modeling task, where a recurrent neural network (RNN) models the conditional distribution over the next element conditio…
In this paper, we develop a Markovian model that deals with the volume offered at the best quote of an electronic order book. The volume of the first limit is a stochastic process whose paths are periodically interrupted and reset to a new value, either by a new limit order submitted inside the spread or by a market or…
MPS selects models for nonstationary time series in real-time.
problem Model selection under nonstationary time series.
method Combines conformal inference with model confidence sets.
result Reliably identifies optimal models under nonstationarity.
Model predicts activist fund targets with 78.2% accuracy.
problem Predicting activist fund targets to mitigate risks and inform investments.
method Evaluated 123 model configurations using machine learning techniques.
result Best model achieved AUC-ROC of 0.782.
Research improves federated text models for next word prediction.
problem Training models on distributed devices efficiently and effectively.
method Employ transfer learning in federated learning for next word prediction.
result Enhancements to current baselines with pretrained embeddings and whole model pretraining.
Motivated by the task of hyperparameter optimization, we introduce the non-stochastic best-arm identification problem. Within the multi-armed bandit literature, the cumulative regret objective enjoys algorithms and analyses for both the non-stochastic and stochastic settings while to the best of our knowledge, the best…
A new method improves language model extrapolation without changing training.
problem Improving language model performance on sequences not seen during training.
method Perturbation-based training procedure for language models.
result The method consistently improves out-of-support prediction.
Improved theoretical guarantees for Top Two algorithms.
problem Theoretical support for best arm identification with bounded distributions.
method General analysis of Top Two methods, identifying desirable properties and replacing sampling step.
result Theoretical support for Top Two algorithms with bounded distributions.
Nontrivial connectivity has allowed the training of very deep networks by addressing the problem of vanishing gradients and offering a more efficient method of reusing parameters. In this paper we make a comparison between residual networks, densely-connected networks and highway networks on an image classification tas…
This paper examines how different decoding algorithms for LLMs align with various goals.
problem Consistency of decoding algorithms with different goals in LLMs.
method Analysis of greedy, lookahead, random sampling, and temperature-scaled random sampling algorithms.
result Random sampling is consistent with the true probability distribution, but other goals require optimal algorithms for specific probability distributions.
Predicts user next location using CDR data.
problem Predicting user's next location using CDR data.
method Data processing framework with deep neural network model.
result Error of prediction decreased from 74% to 55%.
High-dimensional data acquired from biological experiments such as next generation sequencing are subject to a number of confounding effects. These effects include both technical effects, such as variation across batches from instrument noise or sample processing, or institution-specific differences in sample acquisiti…
Despite incredible recent advances in machine learning, building machine learning applications remains prohibitively time-consuming and expensive for all but the best-trained, best-funded engineering organizations. This expense comes not from a need for new and improved statistical models but instead from a lack of sys…
Improved knowledge gradient (iKG) outperforms the original KG algorithm in best arm identification problems.
problem Best arm identification (BAI) problem with limitations of the original KG algorithm.
method Follows the one-step look ahead of KG but chooses the measurement that maximizes the probability of selecting the best arm.
result Improved knowledge gradient (iKG) is asymptotically optimal and easier to extend to variant BAI problems.
SANST uses self-attentive networks with spatial and temporal embeddings for better POI recommendations.
problem Next point-of-interest (POI) recommendation for users based on their history.
method SANST incorporates spatio-temporal patterns into self-attentive networks.
result SANST outperforms state-of-the-art models by up to 13.65% in nDCG@10.
Two methods for model adaptation compared; fine-tuning outperforms Best-of-N in realizable settings.
problem Comparing methods for adapting large language models to new tasks.
method Supervised fine-tuning vs. Best-of-N approach.
result Supervised fine-tuning outperforms Best-of-N in realizable settings.
The paper proposes an algorithm to enumerate K best models with distinct support vectors for SVM.
problem Finding multiple models with distinct support vectors for non-standard machine learning applications.
method A K-best model enumeration algorithm for SVM that efficiently finds models with distinct support vectors in the dual SVM problem.
result The algorithm efficiently finds the next best model with small latency, useful for interactive examination of requirements.
Regularization is an important component of predictive model building. The hybrid bootstrap is a regularization technique that functions similarly to dropout except that features are resampled from other training points rather than replaced with zeros. We show that the hybrid bootstrap offers superior performance to dr…
Proposes a method to improve surrogate modeling and design optimization using latent variables.
problem Improving efficiency in multi-fidelity adaptive sampling without hierarchical assumptions.
method A framework using a latent variable Gaussian process to capture correlations between different fidelity models and optimize adaptive sampling.
result Demonstrates superior performance in convergence rate and robustness compared to existing methods.
The economy globalization measure problem is discussed. Four macroeconomic indices of twenty among the "richest" countries are examined. Four types of "distances" are calculated.Two types of networks are next constructed for each distance measure definition. It is shown that the globalization process can be best charac…
PNNs improve personalized healthcare policies using mixed integer programming.
problem Learning treatment policies for patients with limited data.
method Prescriptive networks (PNNs) trained with mixed integer programming.
result PNNs outperform existing methods in reducing peak blood pressure.
MoE-F combines LLMs online for better time-series prediction.
problem Combining multiple LLMs for online time-series prediction.
method Time-adaptive stochastic filtering techniques to combine experts.
result MoE-F achieves 17% absolute and 48.5% relative F1 measure improvement.
Two algorithms optimize assortment selection for user choices in unknown MNL models.
problem Sequential assortment selection with unknown multinomial logit parameters.
method Upper confidence bound algorithms for MNL contextual bandits.
result Optimal regret bounds for assortment selection problems.
CSL selects best model from library based on covariates.
problem Selecting the best model from a library based on covariates.
method Uses meta learning and cross-validation to find a local minimum.
result Converges at a rate faster than Op(n−1/4) and offers extensive empirical evidence. The paper proves Simon's conjecture for closed Einstein spaces, setting eigenvalue bounds.
problem Proving Simon's conjecture for closed Einstein spaces.
method Developed a method to prove Simon's conjecture for closed Einstein spaces.
result Proved Simon's conjecture for closed Einstein spaces, setting eigenvalue bounds.
Algorithm identifies best policy in MDPs with adaptive sampling.
problem Best policy identification in discounted MDPs with limited samples.
method Derive lower and upper bounds on sample complexity, design KLB-TS algorithm.
result KLB-TS algorithm achieves nearly-optimal sample allocation.
A new framework for robot block-stacking tasks using causal probabilistic models.
problem Robots fail outside controlled environments due to uncertainty and lack of explicit design for all scenarios.
method Causal probabilistic framework combining causal models and probabilistic representations of noise.
result Robots can perceive, reason about, and explain their environment for block-stacking tasks.
We applied Dirac distribution, Bose-Einstein distribution, and occasionally Boltzmann-Gibbs distribution in order to determine which is optimal for income distribution on a large pool of countries. The best fit to the data was observed in the case of Fermi-Dirac distribution, for which the coefficient of determination …
We develop a method for the evaluation of extreme event statistics associated with nonlinear dynamical systems, using a small number of samples. From an initial dataset of design points, we formulate a sequential strategy that provides the 'next-best' data point (set of parameters) that when evaluated results in improv…
Study dynamic batch learning in high-dimensional sparse linear bandits.
problem Dynamic batch learning in high-dimensional sparse linear contextual bandits under batch constraints.
method Characterized fundamental learning limits via regret lower bound and provided matching upper bound.
result Prescribed an optimal scheme for dynamic batch learning in high-dimensional sparse linear contextual bandits.
Better methods to detect insider threats need new anticipatory analytics to capture risky behavior prior to losing data. In search of the best overall classifier, this work empirically scores 88 machine learning algorithms in 16 major families. We extract risk features from the large CERT dataset, which blends real net…
Study improves U.S. monetary policy forecasting by integrating text and data.
problem Forecasting central bank policy decisions, especially the Fed's rate changes.
method Multi-modal approach combining structured data and unstructured text from Fed communications.
result Hybrid models outperform unimodal baselines, achieving a test AUC of 0.83.
Active learning (AL) aims to enable training high performance classifiers with low annotation cost by predicting which subset of unlabelled instances would be most beneficial to label. The importance of AL has motivated extensive research, proposing a wide variety of manually designed AL algorithms with diverse theoret…
HiPPO-Prophecy models can learn dynamical systems without fine-tuning.
problem Learning dynamical systems in context without fine-tuning parameters.
method Introduced a novel weight construction for SSMs that approximates derivatives of input signals.
result Discrete SSMs can predict the next state of any dynamical system after observing previous states.
Optimal best-arm identification with known number of optimal arms.
problem Identifying the best arm in a multi-armed bandit with multiple optimal arms under fixed confidence.
method Deriving a new information-theoretic lower bound and proposing a modified stopping rule.
result Achieving asymptotic instance-optimality with a new lower bound and new stopping rule.
Simpler neural architecture search method using random architectures and regression.
problem Complex algorithms in neural architecture search.
method Train N random architectures, use them to train a regression model, predict validation accuracies, and select top-K architectures.
result More sample efficient and competitive with complex approaches.
Improved inference-time alignment using Best-of-N and smoothing.
problem Reward overoptimization in Best-of-N (BoN) due to poor proxy reward models.
method Introduced Soft Best-of-N (SBoN) and analyzed its performance through KL divergence and regret analysis.
result Smoothing helps SBoN mitigate reward overoptimization, especially when proxy reward quality is low.
Study reveals AI's spontaneous topic changes in text prediction.
problem AI's inability to spontaneously switch topics like humans.
method Defined topic as Token Priority Graphs (TPGs) and analyzed self-attention models.
result AI can only switch topics if lower-priority tokens outnumber higher-priority ones.
Federated survival analysis outperforms local and centralized training, with RSF offering the best balance of discrimination, calibration, and robustness.
problem Survival analysis models require large, diverse cohorts but are limited by privacy regulations and lack of centralized data.
method Federated learning (FL) is used to train shared models without exchanging raw data.
result FL consistently outperforms local training and approaches, and occasionally exceeds centralized performance.
LinFACT identifies all ε-best arms in linear bandits with near-optimal efficiency.
problem Efficiently identifying multiple optimal candidates in high trial-and-error cost tasks.
method LinFACT algorithm designed for linear bandits, with information-theoretic lower bound and upper bound derivation integration.
result LinFACT achieves instance optimality, matching lower bound up to a logarithmic factor.
We propose a computationally efficient random walk on a convex body which rapidly mixes and closely tracks a time-varying log-concave distribution. We develop general theoretical guarantees on the required number of steps; this number can be calculated on the fly according to the distance from and the shape of the next…