Ricardo M. Araujo

Towards understanding the role of learning models in the dynamics of the minority game

We report experiments in a boundedly rational evolutionary game, namely the minority game, where agents apply a very simple learning algorithm to discard bad strategies and create new ones. The results show that even such simplified learning model presents qualitative differences from the behavior of the traditional game, where strategies are fixed and cannot be modified or discarded. We show that this result is qualitatively similar to other, more complex, learning approaches. Also, we study how the learning parameters of our model affect the dynamics of the game and we provide experimental evidence of a high dependence between the behavior of the system and the way fitness is attributed as new strategies enter the game.

Classification of Apple Tree Disorders Using Convolutional Neural Networks

This paper studies the use of Convolutional Neural Networks to automatically detect and classify diseases, nutritional deficiencies and damage by herbicides on apple trees from images of their leaves. This task is fundamental to guarantee a high quality of the resulting yields and is currently largely performed by experts in the field, which can severely limit scale and add to costs. By using a novel data set containing labeled examples consisting of 2539 images from 6 known disorders, we show that trained Convolutional Neural Networks are able to match or outperform experts in this task, achieving a 97.3% accuracy on a hold-out set.

The Impact of Centrality on Individual and Collective Performance in Social Problem-Solving Systems

In this paper, we analyze the dependency between centrality and individual performance in socially-inspired problem-solving systems. By means of extensive numerical simulations, we investigate how individual performance in four different models correlate with four different classical centrality measures. Our main result shows that there is a high linear correlation between centrality and individual performance when individuals systematically exploit central positions. In this case, central individuals tend to deviate from the expected majority contribution behavior. Although there is ample evidence about the relevance of centrality in social problem-solving, our work contributes to understand that some measures correlate better with individual performance than others due to individual traits, a position that is gaining strength in recent studies.

Neural-evolutionary learning in a bounded rationality scenario

This paper presents a neural-evolutionary framework for the simulation of market models in a bounded rationality scenario. Each agent involved in the scenario make use of a population of neural networks in order to make a decision, while inductive learning is performed by means of an evolutionary algorithm. We show that good convergence to the game-theoretic equilibrium is reached within certain parameters.

Two Novel Algorithms for High Quality Motion Estimation in High Definition Video Sequences

In this paper, we propose two new algorithms for high quality motion estimation in high definition digital videos. Both algorithms are based on the use of random features that guarantee robustness to avoid dropping into a local-minimum. The first algorithm was developed from a simple two stage approach where a random stage is complemented by a greedy stage in a very simple fashion. The second algorithm is based on a more refined class of algorithms called Memetic Network Algorithms where each instance of the search may exchange information with its neighbour instances according to some rules that control the information flow. The proposed algorithms were implemented and tested exclusively with high definition sequences against well known fast algorithms like Diamond Search and Three Step Search. The results show that our algorithms can outperform other algorithms in quality yielding an increment in complexity that may be amortized if resources for a parallel execution are available. Additionally, we provide further evidence that fast algorithms do not perform well in high definition.

On the Effects of Network Structure in Population-Based Optimization

Memetic networks are a new class of population-based optimization algorithms that makes use of an underlying network to structure information flow between individuals representing points in the search space. Its main characteristic is the possibility of aggregating several solutions in order to compose new ones and the use of an explicit network to aid search. Algorithms from this class can be used to relate network properties to search performance in optimization tasks. We propose and report on algorithms applied to several benchmark optimization problems. We further show how some network properties - in particular, the existence of hubs - can influence the algorithms performance.

Full Body Person Identification Using the Kinect Sensor

Identifying individuals using biometric data is an important task in surveillance, authentication and even entertainment. This task is more challenging when required to be performed without physical contact and at a distance. Analyzing video footages from individuals for patterns is an active area of research aiming at fulfilling this goal. We describe results on classifiers trained to identify individuals from data collected from 140 subjects walking in front of a Microsoft Kinect sensor, which allows tracking 3D points representing a subjects skeleton. From this data we extract anthropometric and gait attributes to be used by the classifiers. We show that anthropometric features are more important than gait features but using both allows for higher accuracies. Additionally, we explore how different numbers of subjects and numbers of available examples affect accuracy, providing evidences on how effective the proposed methodology can be in different scenarios.

Collaboration Emergence in Social Networks with Informational Natural Selection

Social collaboration can benefit individuals by avoiding efforts and risks inherent of trial-and-error learning. However, social collaboration may demand considerable effort and time. We present a new model of social collaboration based on Informational Natural Selection in order to investigate social problem-solving. We performed a set of social network experiments in which individuals solved problems in a virtual environment. Results show that collaboration can be viewed as a complex systems emergence promoted by a agents behavior that results from Informational Natural Selection.

Assessing the Performance of Convolutional Neural Networks on Classifying Disorders in Apple Tree Leaves

This paper evaluates the deep learning architecture AlexNet applied to the diagnosis of disorders from leaf images using a recent dataset containing five apple tree disorders. It extends previous work by providing a more extensive testing and a dataset validation by using visualization methods. We show that previous results likely overestimate general accuracy, but that the model is able to learn relevant features from the images.

Investigating Crime Rate Prediction Using Street-Level Images and Siamese Convolutional Neural Networks

The analysis of the environment for crime prediction is based on the premise that criminal behavior is influenced by the nature of the environment in which occurs. Street-level images are the closest digital depiction available of the urban environment, in which most street crimes take place. This work proposes a crime rate prediction model that uses street-level images to classify street crimes into low or high crime rate levels. For that, we use a 4-Cardinal Siamese Convolution Neural Network (4-CSCNN) and train and test our analytic model in two regions of Rio de Janeiro, Brazil, that showed high street crime concentrations between the years of 2007 and 2016. With this preliminary experiment, we investigate the use of convolutional neural networks (CNN) for the task of crime rating through visual scene analysis and found possibilities towards automatic crime rate predictions using CNN models.