Jorge de la Calleja

Automatic control for laboratory sterilization process based on arduino hardware

In this paper It is introduced an automated control system for the sterilization process of biological material. The proposed system is composed of four modules: flame detection, pressure sensor, Gas Control Module and CPU Module. This System is based on Arduino Board. Arduino is an Open Hardware platform that allows a fast prototype development, Arduino microcontroller is programmed on C++. making development and tests not a difficult part of overall work.

Automated Classification of Galaxy Images

In this paper we present an experimental study of the performance of three machine learning algorithms applied to the difficult problem of galaxy classification. We use the Naive Bayes classifier, the rule-induction algorithm C4.5 and a recently introduced classifier named random forest (RF). We first employ image processing to standardize the images, eliminating the effects of orientation and scale, then perform principal component analysis to reduce the dimensionality of the data, and finally, classify the galaxy images. Our experiments show that RF obtains the best results considering three, five and seven galaxy types.

Image-based classification of diabetic retinopathy using machine learning

In this paper we present experimental results of an automated method for image-based classification of diabetic retinopathy. The method is divided into three stages: image processing, feature extraction and image classification. In the first stage we have used two image processing techniques in order to enhance their features. Then, the second stage reduces the dimensionality of the images and finds features using the statistical method of principal component analysis. Finally, in the third stage the images are classified using machine learning algorithms, particularly, the naive Bayes classifier, neural networks, k-nearest neighbors and support vector machines. In our experimental study we classify two types of retinopathy: non-proliferative and proliferative. Preliminary results show that k-nearest neighbors obtained the best result with 68.7% using f-measure as metric, for a data set of 151 images with different resolutions.

LBP and Machine Learning for Diabetic Retinopathy Detection

Diabetic retinopathy is a chronic progressive eye disease associated to a group of eye problems as a complication of diabetes. This disease may cause severe vision loss or even blindness. Specialists analyze fundus images in order to diagnostic it and to give specific treatments. Fundus images are photographs taken of the retina using a retinal camera, this is a noninvasive medical procedure that provides a way to analyze the retina in patients with diabetes. The correct classification of these images depends on the ability and experience of specialists, and also the quality of the images. In this paper we present a method for diabetic retinopathy detection. This method is divided into two stages: in the first one, we have used local binary patterns (LBP) to extract local features, while in the second stage, we have applied artificial neural networks, random forest and support vector machines for the detection task. Preliminary results show that random forest was the best classifier with 97.46% of accuracy, using a data set of 71 images.

Associating students and teachers for tutoring in higher education using clustering and data mining

Tutoring is part of the teaching–learning process; this is considered a complementary strategy to support the development of integral and competent professionals. When teachers deal with large groups of students such as in digital learning environments, tutoring becomes a time‐consuming and difficult task that can cause distraction and overload. This paper presents an experimental study to associate students and teachers for tutoring according to their skills and affinities using the clustering methods of k‐means, expectation maximization, and farthest first. The study harvests data of 1,199 university students and 35 teachers. The results reached 100% of compatibility between clusters using expectation maximization and farthest first.

Machine learning from imbalanced data sets for astronomical object classification

In this paper we present an experimental study of machine learning from imbalanced data sets applied to the difficult problem of astronomical object classification in multi-spectral wide-field images. The imbalanced data set problem is very common in several domains, and occurs when there are many more examples of some classes than others; therefore, classifiers perform poorly on these data sets. In order to improve the performance of machine learning algorithms over minority class examples, we propose to create new instances using a modification of the well-known SMOTE technique, but only of those misclassified examples given by an ensemble of classifiers. Our preliminary experimental results show that the proposed approach obtain above. 700 using recall, precision and f-measure as metrics for evaluation; using small data sets.

A 3D simulation environment for kinematic task of the PUMA 560 robot

Robots with articulated morphology has acquired major importance because of representing la basic application to scale to highly articulated chains. Hence, this paper is focus to present how manipulate kinematic chains with six degrees of freedom, in particular way the model associated to PUMA 560 robot. Techniques to compute forward and inverse kinematic are described. Besides, a platform to simulate on three dimension space is presented as a important tool to verify the movements programmed on the robot. Interesting results from inverse kinematic applications are shown.

The imbalanced problem in morphological galaxy classification

In this paper we present an experimental study of the performance of six machine learning algorithms applied to morphological galaxy classification. We also address the learning approach from imbalanced data sets, inherent to many real-world applications, such as astronomical data analysis problems. We used two over-sampling techniques: SMOTE and Resampling, and we vary the amount of generated instances for classification. Our experimental results show that the learning method Random Forest with Resampling obtain the best results for three, five and seven galaxy types, with a F-measure about. 99 for all cases.

Key Elements for Motion Planning Algorithms

Planning a collision-free path for a rigid or articulated robot to move from an initial to a final configuration in a static environment is a central problem in robotics and has been extensively addressed over the last. The complexity of the problem is NP-hard (Latombe, 1991). There exist several family sets of variations of the basic problem, that consider flexible robots, and where robots can modify the environment. The problem is well known in other domains, such as planning for graphics and simulation (Koga et al., 1994), planning for virtual prototyping (Chang & Li, 1995), and planning for medical (Tombropoulos et al., 1999) and pharmaceutical (Finn & Kavraki, 1999) applications.

LOD4AIR: A strategy to produce and consume linked open data from OAI-PMH repositories

This paper proposes a strategy to produce and consume linked open data (LOD) from OAI-PMH compliant repositories. The strategy represents an alternative to support management and reuse of machine-readable knowledge, which can be exploited by means of semantic web services. The papers main contribution includes a characterization of tools used to produce LOD, the specification of semantic web services that consumes LOD, and an analysis of our strategys potential benefits for end-users of REMERI, a specific federated network of institutional repositories. The paper also poses the challenges associated with the implementation of the proposed strategy.