Searcharxiv⌕ Search

arXiv subjects

Ostap Okhrin

Publications and source records attributed to Ostap Okhrin.

31 records · Page 2Linked to original sources
arXiv2022

Modified DDPG car-following model with a real-world human driving experience with CARLA simulator

In the autonomous driving field, fusion of human knowledge into Deep Reinforcement Learning (DRL) is often based on the human demonstration recorded in a simulated environment. This limits the generalization and the feasibility of application in real-world traffic. We propose a two-stage DRL method to train a car-following agent, that modifies the policy by leveraging the real-world human driving experience and achieves performance superior to the pure DRL agent. Training a DRL agent is done within CARLA framework with Robot Operating System (ROS). For evaluation, we designed different driving scenarios to compare the proposed two-stage DRL car-following agent with other agents. After extracting the "good" behavior from the human driver, the agent becomes more efficient and reasonable, which makes this autonomous agent more suitable for Human-Robot Interaction (HRI) traffic.

cs.RO↗
arXiv2022

Vessel-following model for inland waterways based on deep reinforcement learning

While deep reinforcement learning (RL) has been increasingly applied in designing car-following models in the last years, this study aims at investigating the feasibility of RL-based vehicle-following for complex vehicle dynamics and strong environmental disturbances. As a use case, we developed an inland waterways vessel-following model based on realistic vessel dynamics, which considers environmental influences, such as varying stream velocity and river profile. We extracted natural vessel behavior from anonymized AIS data to formulate a reward function that reflects a realistic driving style next to comfortable and safe navigation. Aiming at high generalization capabilities, we propose an RL training environment that uses stochastic processes to model leading trajectory and river dynamics. To validate the trained model, we defined different scenarios that have not been seen in training, including realistic vessel-following on the Middle Rhine. Our model demonstrated safe and comfortable driving in all scenarios, proving excellent generalization abilities. Furthermore, traffic oscillations could effectively be dampened by deploying the trained model on a sequence of following vessels.

cs.CE↗
arXiv2022

Vulnerability-CoVaR: Investigating the Crypto-market

This paper proposes an important extension to Conditional Value-at-Risk (CoVaR), the popular systemic risk measure, and investigates its properties on the cryptocurrency market. The proposed Vulnerability-CoVaR (VCoVaR) is defined as the Value-at-Risk (VaR) of a financial system or institution, given that at least one other institution is equal or below its VaR. The VCoVaR relaxes normality assumptions and is estimated via copula. While important theoretical findings of the measure are detailed, the empirical study analyzes how different distressing events of the cryptocurrencies impact the risk level of each other. The results show that Litecoin displays the largest impact on Bitcoin and that each cryptocurrency is significantly affected if an event of joint distress among the remaining market participants occurs. The VCoVaR is shown to capture domino effects better than other CoVaR extensions.

q-fin.GN↗
arXiv2021

Missing Velocity in Dynamic Obstacle Avoidance based on Deep Reinforcement Learning

We introduce a novel approach to dynamic obstacle avoidance based on Deep Reinforcement Learning by defining a traffic type independent environment with variable complexity. Filling a gap in the current literature, we thoroughly investigate the effect of missing velocity information on an agent's performance in obstacle avoidance tasks. This is a crucial issue in practice since several sensors yield only positional information of objects or vehicles. We evaluate frequently-applied approaches in scenarios of partial observability, namely the incorporation of recurrency in the deep neural networks and simple frame-stacking. For our analysis, we rely on state-of-the-art model-free deep RL algorithms. The lack of velocity information is found to significantly impact the performance of an agent. Both approaches - recurrency and frame-stacking - cannot consistently replace missing velocity information in the observation space. However, in simplified scenarios, they can significantly boost performance and stabilize the overall training procedure.

cs.LG↗
arXiv2021

Formulation and validation of a car-following model based on deep reinforcement learning

We propose and validate a novel car following model based on deep reinforcement learning. Our model is trained to maximize externally given reward functions for the free and car-following regimes rather than reproducing existing follower trajectories. The parameters of these reward functions such as desired speed, time gap, or accelerations resemble that of traditional models such as the Intelligent Driver Model (IDM) and allow for explicitly implementing different driving styles. Moreover, they partially lift the black-box nature of conventional neural network models. The model is trained on leading speed profiles governed by a truncated Ornstein-Uhlenbeck process reflecting a realistic leader's kinematics. This allows for arbitrary driving situations and an infinite supply of training data. For various parameterizations of the reward functions, and for a wide variety of artificial and real leader data, the model turned out to be unconditionally string stable, comfortable, and crash-free. String stability has been tested with a platoon of five followers following an artificial and a real leading trajectory. A cross-comparison with the IDM calibrated to the goodness-of-fit of the relative gaps showed a higher reward compared to the traditional model and a better goodness-of-fit.

cs.LG↗
arXiv2020

Outer power transformations of hierarchical Archimedean copulas: Construction, sampling and estimation

A large number of commonly used parametric Archimedean copula (AC) families are restricted to a single parameter, connected to a concordance measure such as Kendall's tau. This often leads to poor statistical fits, particularly in the joint tails, and can sometimes even limit the ability to model concordance or tail dependence mathematically. This work suggests outer power (OP) transformations of Archimedean generators to overcome these limitations. The copulas generated by OP-transformed generators can, for example, allow one to capture both a given concordance measure and a tail dependence coefficient simultaneously. For exchangeable OP-transformed ACs, a formula for computing tail dependence coefficients is obtained, as well as two feasible OP AC estimators are proposed and their properties studied by simulation. For hierarchical extensions of OP-transformed ACs, a new construction principle, efficient sampling and parameter estimation are addressed. By simulation, convergence rate and standard errors of the proposed estimator are studied. Excellent tail fitting capabilities of OP-transformed hierarchical AC models are demonstrated in a risk management application. The results show that the OP transformation is able to improve the statistical fit of exchangeable ACs, particularly of those that cannot capture upper tail dependence or strong concordance, as well as the statistical fit of hierarchical ACs, especially in terms of tail dependence and higher dimensions. Given how comparably simple it is to include OP transformations into existing exchangeable and hierarchical AC models, this transformation provides an attractive trade-off between computational effort and statistical improvement.

stat.ME↗
arXiv2019

Infinitely Stochastic Micro Forecasting

Forecasting costs is now a front burner in empirical economics. We propose an unconventional tool for stochastic prediction of future expenses based on the individual (micro) developments of recorded events. Consider a firm, enterprise, institution, or state, which possesses knowledge about particular historical events. For each event, there is a series of several related subevents: payments or losses spread over time, which all leads to an infinitely stochastic process at the end. Nevertheless, the issue is that some already occurred events do not have to be necessarily reported. The aim lies in forecasting future subevent flows coming from already reported, occurred but not reported, and yet not occurred events. Our methodology is illustrated on quantitative risk assessment, however, it can be applied to other areas such as startups, epidemics, war damages, advertising and commercials, digital payments, or drug prescription as manifested in the paper. As a theoretical contribution, inference for infinitely stochastic processes is developed. In particular, a non-homogeneous Poisson process with non-homogeneous Poisson processes as marks is used, which includes for instance the Cox process as a special case.

econ.EM↗
arXiv2018

Optimal Shrinkage Estimator for High-Dimensional Mean Vector

In this paper we derive the optimal linear shrinkage estimator for the high-dimensional mean vector using random matrix theory. The results are obtained under the assumption that both the dimension $p$ and the sample size $n$ tend to infinity in such a way that $p/n \to c\in(0,\infty)$. Under weak conditions imposed on the underlying data generating mechanism, we find the asymptotic equivalents to the optimal shrinkage intensities and estimate them consistently. The proposed nonparametric estimator for the high-dimensional mean vector has a simple structure and is proven to minimize asymptotically, with probability $1$, the quadratic loss when $c\in(0,1)$. When $c\in(1, \infty)$ we modify the estimator by using a feasible estimator for the precision covariance matrix. To this end, an exhaustive simulation study and an application to real data are provided where the proposed estimator is compared with known benchmarks from the literature. It turns out that the existing estimators of the mean vector, including the new proposal, converge to the sample mean vector when the true mean vector has an unbounded Euclidean norm.

math.ST↗
arXiv2018

Dynamic and granular loss reserving with copulae

An intensive research sprang up for stochastic methods in insurance during the past years. To meet all future claims rising from policies, it is requisite to quantify the outstanding loss liabilities. Loss reserving methods based on aggregated data from run-off triangles are predominantly used to calculate the claims reserves. Conventional reserving techniques have some disadvantages: loss of information from the policy and the claim's development due to the aggregation, zero or negative cells in the triangle; usually small number of observations in the triangle; only few observations for recent accident years; and sensitivity to the most recent paid claims. To overcome these dilemmas, granular loss reserving methods for individual claim-by-claim data will be derived. Reserves' estimation is a crucial part of the risk valuation process, which is now a front burner in economics. Since there is a growing demand for prediction of total reserves for different types of claims or even multiple lines of business, a time-varying copula framework for granular reserving will be established.

econ.EM↗
arXiv2017

Index of Environmental Awareness in Russia - MIMIC Approaches for Different Economic Situations

The paper addresses the issue of environmental awareness in the regions of the Russian Federation. Russia provides an important study area to investigate the relationship between economic development and environmental consciousness in general. This paper introduces an index of environmental awareness, which is derived as a latent variable from various categories of search entries in Yandex, the prominent Russian search engine, during two periods in years 2014 and 2015. These indicators are presumably dependent on certain causes, which are also integrated into the model. The resulting Multiple Indicators-Multiple Causes model allows to estimate the proposed index of environmental awareness and to rank the Russian regions for each period. Comparing the results of 2014 and 2015 is especially interesting, because of the RUR devaluation at the end of 2014. In additional, we answer the question of the existence of an Environmental Kuznets Curve with respect to environmental understanding.

stat.AP↗
arXiv2013

Conditional Least Squares and Copulae in Claims Reserving for a Single Line of Business

One of the main goals in non-life insurance is to estimate the claims reserve distribution. A generalized time series model, that allows for modeling the conditional mean and variance of the claim amounts, is proposed for the claims development. On contrary to the classical stochastic reserving techniques, the number of model parameters does not depend on the number of development periods, which leads to a more precise forecasting. Moreover, the time series innovations for the consecutive claims are not considered to be independent anymore. Conditional least squares are used for model parameter estimation and consistency of such estimate is proved. Copula approach is used for modeling the dependence structure, which improves the precision of the reserve distribution estimate as well. Real data examples are provided as an illustration of the potential benefits of the presented approach.

stat.AP↗
arXiv2011

Asymptotic behaviour of the S-stopped branching processes with countable state space

he starting process with countable number of types μ(t) generates a stopped branching process ξ(t). The starting process stops, by falling into the nonempty set S. It is assumed, that the starting process is subcritical, indecomposable and noncyclic. It is proved, that the extinction probability converges to the cyclic function with period 1.

math.ST↗