Waldo Hasperué

Some stylized facts of the Bitcoin market

In recent years a new type of tradable assets appeared, generically known as cryptocurrencies. Among them, the most widespread is Bitcoin. Given its novelty, this paper investigates some statistical properties of the Bitcoin market. This study compares Bitcoin and standard currencies dynamics and focuses on the analysis of returns at different time scales. We test the presence of long memory in return time series from 2011 to 2017, using transaction data from one Bitcoin platform. We compute the Hurst exponent by means of the Detrended Fluctuation Analysis method, using a sliding window in order to measure long range dependence. We detect that Hurst exponents changes significantly during the first years of existence of Bitcoin, tending to stabilize in recent times. Additionally, multiscale analysis shows a similar behavior of the Hurst exponent, implying a self-similar process.

Geographical patterns of Triatominae (Heteroptera: Reduviidae) richness and distribution in the Western Hemisphere

Broad‐scale spatial patterns in species richness have been widely investigated with spatial statistics tools in the past few years. The primary goal of these investigations has been to understand the ecological and evolutionary processes underlying such patterns. Nevertheless, most of the current (climate) explanations for these patterns actually rely on the geographical range limits of species, so that a better understanding of such processes may be achieved by coupling richness and distribution (niche) models. We analysed the geographical ranges and richness patterns for 115 triatomine species in the Neotropics, modelled as a function of 12 environmental variables expressing alternative hypotheses that have been used to explain richness gradients. These analyses were based on spatial [spatial eigenvector mapping (SEVM)] and non‐spatial ordinary least‐squares multiple regression models. The geographical ranges of species were also individually analysed using a general linear model (GLM). The coefficients of the regression models for richness and distribution were then compared. Spatial analyses revealed that the unique contributions of spatial eigenvectors and environmental variables to richness were, respectively, equal to 24.2% and 12.2%, with high coefficient values for temperature, actual evapotranspiration, and seasonality. Similar results were obtained using a GLM, and the mean GLM coefficients had a relatively high correlation with those obtained with SEVM (r = 0.586; P < 0.05). Our analyses show that the drivers of Neotropical Triatominae richness and of its species ranges show a high correlation, although the differences among the drivers may be important for understanding the emergent properties (historical processes and species‐specific environmental drivers) that explain richness patterns. Moreover, although our analyses identified an important role for temperature and temperature seasonality in explaining both species richness and distributions, other spatially structured environmental variables and historical factors may explain a large part of the variation in diversity patterns.

Rule Extraction on Numeric Datasets Using Hyper-rectangles

When there is a need to understand the data stored in a database, one of the main requirements is being able to extract knowledge in the form of rules. Classification strategies allow extracting rules almost naturally. In this paper, a new classification strategy is presented that uses hyper-rectangles as data descriptors to achieve a model that allows extracting knowledge in the form of classification rules. The participation of an expert for training the model is discussed. Finally, the results obtained using the databases from the UCI repository are presented and compared with other existing classification models, showing that the algorithm presented requires less computational resources and achieves the same accuracy level and number of extracted rules.

Obtaining a fuzzy classification rule system from a non-supervised clustering

The fuzzy classification systems have been broadly used to solve control and decision-making problem. However, its design is complex, even when having a human expert assistance. This paper presents a new strategy capable of automatically defining the corresponding Fuzzy Classification Rule System from a non-supervised clustering of the available data. Its application to three data sets of the UCI repository has given quite satisfactory results.

GAAP. genetic algorithm with auxiliary populations applied to continuous optimization problems

Genetic algorithms have been used successfully to solve continuous optimization problems. However, an early convergence to low-quality solutions is one of the most common difficulties encountered when using these strategies. In this paper, a method that combines multiple auxiliary populations with the main population of the algorithm is proposed. The role of the auxiliary populations is dual: to prevent or hinder the early convergence to local suboptimal solutions, and to provide a local search mechanism for a greater exploitation of the most promising regions within the search space.

Classification Rules Obtained from Evidence Accumulation

This paper presents a machine learning approach applicable to Data Mining based on obtaining classification rules. It proposes a strategy to obtain classification rules from clusters resulting from a co-association matrix. Such matrix is obtained from the combination of different clustering methods applied to input data, and it has been selected by its results robustness. The proposed method has been applied to two sets of data obtained from the UCl repository with really successful results. The results obtained in the classification have been compared to other existing methods showing the new proposed method superiority.