User profiles for Y. Nevmyvaka
yuriy nevmyvakaManaging Director, ML Research, Morgan Stanley Verified email at morganstanley.com Cited by 982 |
Reinforcement learning for optimized trade execution
Y Nevmyvaka, Y Feng, M Kearns - Proceedings of the 23rd international …, 2006 - dl.acm.org
… the immediate (1-step) cost of taking action a in state x; y is the new state where we end up
after taking a in x; n is the number of times we have tried a in x, and p is the action taken in y. …
after taking a in x; n is the number of times we have tried a in x, and p is the action taken in y. …
Provably convergent Schrödinger bridge with applications to probabilistic time series imputation
The Schrödinger bridge problem (SBP) is gaining increasing attention in generative modeling
and showing promising potential even in comparison with the score-based generative …
and showing promising potential even in comparison with the score-based generative …
Empirical limitations on high frequency trading profitability
Addressing the ongoing examination of high-frequency trading practices in financial markets,
we report the results of an extensive empirical study estimating the maximum possible …
we report the results of an extensive empirical study estimating the maximum possible …
Lag-llama: Towards foundation models for time series forecasting
Over the past years, foundation models have caused a paradigm shift in machine learning
due to their unprecedented capabilities for zero-shot and few-shot generalization. However, …
due to their unprecedented capabilities for zero-shot and few-shot generalization. However, …
[PDF][PDF] Machine learning for market microstructure and high frequency trading
M Kearns, Y Nevmyvaka - High Frequency Trading: New Realities for …, 2013 - c.mql5.com
In this chapter, we overview the uses of machine learning for high frequency trading and
market microstructure data and problems. Machine learning is a vibrant subfield of computer …
market microstructure data and problems. Machine learning is a vibrant subfield of computer …
Modeling temporal data as continuous functions with stochastic process diffusion
Temporal data such as time series can be viewed as discretized measurements of the
underlying function. To build a generative model for such data we have to model the stochastic …
underlying function. To build a generative model for such data we have to model the stochastic …
Censored exploration and the dark pool problem
Dark pools are a recent type of stock exchange in which information about outstanding orders
is deliberately hidden in order to minimize the market impact of large-volume trades. The …
is deliberately hidden in order to minimize the market impact of large-volume trades. The …
Risk bounds on aleatoric uncertainty recovery
… Without loss of generality, consider a regression task in the feature space X (eg, Rd) with
response Y taking values in Y … ) ∼D, and allows for the general form of heteroscedasticity of Y . …
response Y taking values in Y … ) ∼D, and allows for the general form of heteroscedasticity of Y . …
In-or out-of-distribution detection via dual divergence estimation
…, A Schneider, Y Nevmyvaka - Uncertainty in …, 2023 - proceedings.mlr.press
Detecting out-of-distribution (OOD) samples is a problem of practical importance for a reliable
use of deep neural networks (DNNs) in production settings. The corollary to this problem is …
use of deep neural networks (DNNs) in production settings. The corollary to this problem is …
Modeling temporal data as continuous functions with process diffusion
Temporal data like time series are often observed at irregular intervals which is a challenging
setting for the existing machine learning methods. To tackle this problem, we view such …
setting for the existing machine learning methods. To tackle this problem, we view such …