This paper studies the informativeness of features extracted from a limit
order book data, to classify market data vector into the label (buy/idle) by
using the Long short-term memory (LSTM) network. New technical indicators
based on the support/resistance zones are introduced to enrich the set of
features. We evaluate whether the performance of the LSTM network model is
improved when we select features with respect to the newly proposed
methods. Moreover, we employ multicriteria optimization to perform adequate
feature selection among the proposed approaches, with respect to precision,
recall, and F? score. Seven variations of approaches to select features are
proposed and the best is selected by incorporation of multicriteria
optimization.
This paper considers the Risk-constrained Cash-in-Transit Vehicle Routing Problem (RCTVRP), a variant of the vehicle routing problem which takes into account risk factors of the routes. In this problem, the risk constraints are set by using a risk threshold T on each route and thus, the routes with risk larger than T are forbidden. The main idea of this paper is to use the possibility of being robbed along each route, instead of just allowing solutions with routes that satisfy the risk constraints. We develop a new fuzzy version of the RCTVRP, called FRCTVRP, which considers the value of the risk index of each route and the solutions with lower values of risk indexes on their routes are considered as better. In order to achieve that, fuzzy numbers are incorporated into the new formulation. Moreover, two mixed integer program formulations of the FRCTVRP are developed in the paper. The introduced FRCTVRP is compared with the classical RCTVRP from the literature on an adequate example and the advantage of the newly proposed FRCTVRP is demonstrated. Computational experiments are performed and the comparison given in the paper shows that our approach leads to safer routes.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.