Incremental filter and wrapper approaches for feature discretization

Discrete data representations are necessary, or at least convenient, in many machine learning problems. While feature selection (FS) techniques aim at finding relevant subsets of features, the goal of feature discretization (FD) is to find concise (quantized) data representations, adequate for the learning task at hand. In this paper, we propose two incremental methods for FD. The first method belongs to the filter family, in which the quality of the discretization is assessed by a (supervised or unsupervised) relevance criterion. The second method is a wrapper, where discretized features are assessed using a classifier. Both methods can be coupled with any static (unsupervised or supervised) discretization procedure and can be used to perform FS as pre-processing or post-processing stages. The proposed methods attain efficient representations suitable for binary and multi-class problems with different types of data, being competitive with existing methods. Moreover, using well-known FS methods with the features discretized by our techniques leads to better accuracy than with the features discretized by other methods or with the original features. (C) 2013 Elsevier B.V. All rights reserved.

Mapping the Pasture Steppe in Bayankhongor, Mongolia: comparison of classification methods, using Landsat-8 and geophysical data

Grasslands in semi-arid regions, like Mongolian steppes, are facing desertification and degradation processes, due to climate change. Mongolia’s main economic activity consists on an extensive livestock production and, therefore, it is a concerning matter for the decision makers. Remote sensing and Geographic Information Systems provide the tools for advanced ecosystem management and have been widely used for monitoring and management of pasture resources. This study investigates which is the higher thematic detail that is possible to achieve through remote sensing, to map the steppe vegetation, using medium resolution earth observation imagery in three districts (soums) of Mongolia: Dzag, Buutsagaan and Khureemaral. After considering different thematic levels of detail for classifying the steppe vegetation, the existent pasture types within the steppe were chosen to be mapped. In order to investigate which combination of data sets yields the best results and which classification algorithm is more suitable for incorporating these data sets, a comparison between different classification methods were tested for the study area. Sixteen classifications were performed using different combinations of estimators, Landsat-8 (spectral bands and Landsat-8 NDVI-derived) and geophysical data (elevation, mean annual precipitation and mean annual temperature) using two classification algorithms, maximum likelihood and decision tree. Results showed that the best performing model was the one that incorporated Landsat-8 bands with mean annual precipitation and mean annual temperature (Model 13), using the decision tree. For maximum likelihood, the model that incorporated Landsat-8 bands with mean annual precipitation (Model 5) and the one that incorporated Landsat-8 bands with mean annual precipitation and mean annual temperature (Model 13), achieved the higher accuracies for this algorithm. The decision tree models consistently outperformed the maximum likelihood ones.

Brain Tissue Classification from Multispectral MRI by Wavelet based Principal Component Analysis

In this paper, we propose a multispectral analysis system using wavelet based Principal Component Analysis (PCA), to improve the brain tissue classification from MRI images. Global transforms like PCA often neglects significant small abnormality details, while dealing with a massive amount of multispectral data. In order to resolve this issue, input dataset is expanded by detail coefficients from multisignal wavelet analysis. Then, PCA is applied on the new dataset to perform feature analysis. Finally, an unsupervised classification with Fuzzy C-Means clustering algorithm is used to measure the improvement in reproducibility and accuracy of the results. A detailed comparative analysis of classified tissues with those from conventional PCA is also carried out. Proposed method yielded good improvement in classification of small abnormalities with high sensitivity/accuracy values, 98.9/98.3, for clinical analysis. Experimental results from synthetic and clinical data recommend the new method as a promising approach in brain tissue analysis.

Image segmentation integrating colour, texture and boundary information

La tesis se centra en la Visión por Computador y, más concretamente, en la segmentación de imágenes, la cual es una de las etapas básicas en el análisis de imágenes y consiste en la división de la imagen en un conjunto de regiones visualmente distintas y uniformes considerando su intensidad, color o textura. Se propone una estrategia basada en el uso complementario de la información de región y de frontera durante el proceso de segmentación, integración que permite paliar algunos de los problemas básicos de la segmentación tradicional. La información de frontera permite inicialmente identificar el número de regiones presentes en la imagen y colocar en el interior de cada una de ellas una semilla, con el objetivo de modelar estadísticamente las características de las regiones y definir de esta forma la información de región. Esta información, conjuntamente con la información de frontera, es utilizada en la definición de una función de energía que expresa las propiedades requeridas a la segmentación deseada: uniformidad en el interior de las regiones y contraste con las regiones vecinas en los límites. Un conjunto de regiones activas inician entonces su crecimiento, compitiendo por los píxeles de la imagen, con el objetivo de optimizar la función de energía o, en otras palabras, encontrar la segmentación que mejor se adecua a los requerimientos exprsados en dicha función. Finalmente, todo esta proceso ha sido considerado en una estructura piramidal, lo que nos permite refinar progresivamente el resultado de la segmentación y mejorar su coste computacional. La estrategia ha sido extendida al problema de segmentación de texturas, lo que implica algunas consideraciones básicas como el modelaje de las regiones a partir de un conjunto de características de textura y la extracción de la información de frontera cuando la textura es presente en la imagen. Finalmente, se ha llevado a cabo la extensión a la segmentación de imágenes teniendo en cuenta las propiedades de color y textura. En este sentido, el uso conjunto de técnicas no-paramétricas de estimación de la función de densidad para la descripción del color, y de características textuales basadas en la matriz de co-ocurrencia, ha sido propuesto para modelar adecuadamente y de forma completa las regiones de la imagen. La propuesta ha sido evaluada de forma objetiva y comparada con distintas técnicas de integración utilizando imágenes sintéticas. Además, se han incluido experimentos con imágenes reales con resultados muy positivos.

Case study: object and ground point separation on a river bank using airborne based LIDAR data

In the past decade, airborne based LIght Detection And Ranging (LIDAR) has been recognised by both the commercial and public sectors as a reliable and accurate source for land surveying in environmental, engineering and civil applications. Commonly, the first task to investigate LIDAR point clouds is to separate ground and object points. Skewness Balancing has been proven to be an efficient non-parametric unsupervised classification algorithm to address this challenge. Initially developed for moderate terrain, this algorithm needs to be adapted to handle sloped terrain. This paper addresses the difficulty of object and ground point separation in LIDAR data in hilly terrain. A case study on a diverse LIDAR data set in terms of data provider, resolution and LIDAR echo has been carried out. Several sites in urban and rural areas with man-made structure and vegetation in moderate and hilly terrain have been investigated and three categories have been identified. A deeper investigation on an urban scene with a river bank has been selected to extend the existing algorithm. The results show that an iterative use of Skewness Balancing is suitable for sloped terrain.

Threshold-free object and ground point separation in LIDAR data

Light Detection And Ranging (LIDAR) is an important modality in terrain and land surveying for many environmental, engineering and civil applications. This paper presents the framework for a recently developed unsupervised classification algorithm called Skewness Balancing for object and ground point separation in airborne LIDAR data. The main advantages of the algorithm are threshold-freedom and independence from LIDAR data format and resolution, while preserving object and terrain details. The framework for Skewness Balancing has been built in this contribution with a prediction model in which unknown LIDAR tiles can be categorised as “hilly” or “moderate” terrains. Accuracy assessment of the model is carried out using cross-validation with an overall accuracy of 95%. An extension to the algorithm is developed to address the overclassification issue for hilly terrain. For moderate terrain, the results show that from the classified tiles detached objects (buildings and vegetation) and attached objects (bridges and motorway junctions) are separated from bare earth (ground, roads and yards) which makes Skewness Balancing ideal to be integrated into geographic information system (GIS) software packages.

Aplicação do Sensoriamento Remoto e do Sistema de Informações Geográficas na Detecção de Manchas de Óleo na Região do Pólo de Exploração de Guamaré, R.N.

Objective to establish a methodology for the oil spill monitoring on the sea surface, located at the Submerged Exploration Area of the Polo Region of Guamaré, in the State of Rio Grande do Norte, using orbital images of Synthetic Aperture Radar (SAR integrated with meteoceanographycs products. This methodology was applied in the following stages: (1) the creation of a base map of the Exploration Area; (2) the processing of NOAA/AVHRR and ERS-2 images for generation of meteoceanographycs products; (3) the processing of RADARSAT-1 images for monitoring of oil spills; (4) the integration of RADARSAT-1 images with NOAA/AVHRR and ERS-2 image products; and (5) the structuring of a data base. The Integration of RADARSAT-1 image of the Potiguar Basin of day 21.05.99 with the base map of the Exploration Area of the Polo Region of Guamaré for the identification of the probable sources of the oil spots, was used successfully in the detention of the probable spot of oil detected next to the exit to the submarine emissary in the Exploration Area of the Polo Region of Guamaré. To support the integration of RADARSAT-1 images with NOAA/AVHRR and ERS-2 image products, a methodology was developed for the classification of oil spills identified by RADARSAT-1 images. For this, the following algorithms of classification not supervised were tested: K-means, Fuzzy k-means and Isodata. These algorithms are part of the PCI Geomatics software, which was used for the filtering of RADARSAT-1 images. For validation of the results, the oil spills submitted to the unsupervised classification were compared to the results of the Semivariogram Textural Classifier (STC). The mentioned classifier was developed especially for oil spill classification purposes and requires PCI software for the whole processing of RADARSAT-1 images. After all, the results of the classifications were analyzed through Visual Analysis; Calculation of Proportionality of Largeness and Analysis Statistics. Amongst the three algorithms of classifications tested, it was noted that there were no significant alterations in relation to the spills classified with the STC, in all of the analyses taken into consideration. Therefore, considering all the procedures, it has been shown that the described methodology can be successfully applied using the unsupervised classifiers tested, resulting in a decrease of time in the identification and classification processing of oil spills, if compared with the utilization of the STC classifier

Land use image classification through optimum-path forest clustering

Land use classification has been paramount in the last years, since we can identify illegal land use and also to monitor deforesting areas. Although one can find several research works in the literature that address this problem, we propose here the land use recognition by means of Optimum-Path Forest Clustering (OPF), which has never been applied to this context up to date. Experiments among Optimum-Path Forest, Mean Shift and K-Means demonstrated the robustness of OPF for automatic land use classification of images obtained by CBERS-2B and Ikonos-2 satellites. © 2011 IEEE.

Geoestística e sensoriamento remoto na classificação de imagens em áreas cultivadas com citros

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Segmentação interativa de imagens utilizando competição e cooperação entre partículas

Pós-graduação em Ciência da Computação - IBILCE

Flash floods and their effects on the development in El-Qaá plain area in South Sinai, Egypt

The arid regions are dominated to a much larger degree than humid regions by major catastrophic events. Although most of Egypt lies within the great hot desert belt; it experiences especially in the north some torrential rainfall, which causes flash floods all over Sinai Peninsula. Flash floods in hot deserts are characterized by high velocity and low duration with a sharp discharge peak. Large sediment loads may be carried by floods threatening fields and settlements in the wadis and even people who are living there. The extreme spottiness of rare heavy rainfall, well known to desert people everywhere, precludes any efficient forecasting. Thus, although the limitation of data still reflects pre-satellite methods, chances of developing a warning system for floods in the desert seem remote. The relatively short flood-to-peak interval, a characteristic of desert floods, presents an additional impediment to the efficient use of warning systems. The present thesis contains introduction and five chapters, chapter one points out the physical settings of the study area. There are the geological settings such as outcrop lithology of the study area and the deposits. The alluvial deposits of Wadi Moreikh had been analyzed using OSL dating to know deposits and palaeoclimatic conditions. The chapter points out as well the stratigraphy and the structure geology containing main faults and folds. In addition, it manifests the pesent climate conditions such as temperature, humidity, wind and evaporation. Besides, it presents type of soils and natural vegetation cover of the study area using unsupervised classification for ETM+ images. Chapter two points out the morphometric analysis of the main basins and their drainage network in the study area. It is divided into three parts: The first part manifests the morphometric analysis of the drainage networks which had been extracted from two main sources, topographic maps and DEM images. Basins and drainage networks are considered as major influencing factors on the flash floods; Most of elements were studied which affect the network such as stream order, bifurcation ratio, stream lengths, stream frequency, drainage density, and drainage patterns. The second part of this chapter shows the morphometric analysis of basins such as area, dimensions, shape and surface. Whereas, the third part points the morphometric analysis of alluvial fans which form most of El-Qaá plain. Chapter three manifests the surface runoff through rainfall and losses analysis. The main subject in this chapter is rainfall which has been studied in detail; it is the main reason for runoff. Therefore, all rainfall characteristics are regarded here such as rainfall types, distribution, rainfall intensity, duration, frequency, and the relationship between rainfall and runoff. While the second part of this chapter concerns with water losses estimation by evaporation and infiltration which are together the main losses with direct effect on the high of runoff. Finally, chapter three points out the factors influencing desert runoff and runoff generation mechanism. Chapter four is concerned with assessment of flood hazard, it is important to estimate runoff and tocreate a map of affected areas. Therefore, the chapter consists of four main parts; first part manifests the runoff estimation, the different methods to estimate runoff and its variables such as runoff coefficient lag time, time of concentration, runoff volume, and frequency analysis of flash flood. While the second part points out the extreme event analysis. The third part shows the map of affected areas for every basin and the flash floods degrees. In this point, it has been depending on the DEM to extract the drainage networks and to determine the main streams which are normally more dangerous than others. Finally, part four presets the risk zone map of total study area which is of high inerest for planning activities. Chapter five as the last chapter concerns with flash flood Hazard mitigation. It consists of three main parts. First flood prediction and the method which can be used to predict and forecast the flood. The second part aims to determine the best methods which can be helpful to mitigate flood hazard in the arid zone and especially the study area. Whereas, the third part points out the development perspective for the study area indicating the suitable places in El-Qaá plain for using in economic activities.

Lesion detection and classification in breast cancer: evaluation of approaches based on morphological features, tracer kinetic modelling and semi-quantitative parameters in MR functional imaging (DCE-MRI)

The diagnosis, grading and classification of tumours has benefited considerably from the development of DCE-MRI which is now essential to the adequate clinical management of many tumour types due to its capability in detecting active angiogenesis. Several strategies have been proposed for DCE-MRI evaluation. Visual inspection of contrast agent concentration curves vs time is a very simple yet operator dependent procedure, therefore more objective approaches have been developed in order to facilitate comparison between studies. In so called model free approaches, descriptive or heuristic information extracted from time series raw data have been used for tissue classification. The main issue concerning these schemes is that they have not a direct interpretation in terms of physiological properties of the tissues. On the other hand, model based investigations typically involve compartmental tracer kinetic modelling and pixel-by-pixel estimation of kinetic parameters via non-linear regression applied on region of interests opportunely selected by the physician. This approach has the advantage to provide parameters directly related to the pathophysiological properties of the tissue such as vessel permeability, local regional blood flow, extraction fraction, concentration gradient between plasma and extravascular-extracellular space. Anyway, nonlinear modelling is computational demanding and the accuracy of the estimates can be affected by the signal-to-noise ratio and by the initial solutions. The principal aim of this thesis is investigate the use of semi-quantitative and quantitative parameters for segmentation and classification of breast lesion. The objectives can be subdivided as follow: describe the principal techniques to evaluate time intensity curve in DCE-MRI with focus on kinetic model proposed in literature; to evaluate the influence in parametrization choice for a classic bi-compartmental kinetic models; to evaluate the performance of a method for simultaneous tracer kinetic modelling and pixel classification; to evaluate performance of machine learning techniques training for segmentation and classification of breast lesion.

Deep learning for computer vision: a comparison between convolutional neural networks and hierarchical temporal memories on object recognition tasks

In recent years, Deep Learning techniques have shown to perform well on a large variety of problems both in Computer Vision and Natural Language Processing, reaching and often surpassing the state of the art on many tasks. The rise of deep learning is also revolutionizing the entire field of Machine Learning and Pattern Recognition pushing forward the concepts of automatic feature extraction and unsupervised learning in general. However, despite the strong success both in science and business, deep learning has its own limitations. It is often questioned if such techniques are only some kind of brute-force statistical approaches and if they can only work in the context of High Performance Computing with tons of data. Another important question is whether they are really biologically inspired, as claimed in certain cases, and if they can scale well in terms of "intelligence". The dissertation is focused on trying to answer these key questions in the context of Computer Vision and, in particular, Object Recognition, a task that has been heavily revolutionized by recent advances in the field. Practically speaking, these answers are based on an exhaustive comparison between two, very different, deep learning techniques on the aforementioned task: Convolutional Neural Network (CNN) and Hierarchical Temporal memory (HTM). They stand for two different approaches and points of view within the big hat of deep learning and are the best choices to understand and point out strengths and weaknesses of each of them. CNN is considered one of the most classic and powerful supervised methods used today in machine learning and pattern recognition, especially in object recognition. CNNs are well received and accepted by the scientific community and are already deployed in large corporation like Google and Facebook for solving face recognition and image auto-tagging problems. HTM, on the other hand, is known as a new emerging paradigm and a new meanly-unsupervised method, that is more biologically inspired. It tries to gain more insights from the computational neuroscience community in order to incorporate concepts like time, context and attention during the learning process which are typical of the human brain. In the end, the thesis is supposed to prove that in certain cases, with a lower quantity of data, HTM can outperform CNN.

Stratification and Prognostic Relevance of Jass's Molecular Classification of Colorectal Cancer

Background: The current proposed model of colorectal tumorigenesis is based primarily on CpG island methylator phenotype (CIMP), microsatellite instability (MSI), KRAS, BRAF, and methylation status of 0-6-Methylguanine DNA Methyltransferase (MGMT) and classifies tumors into five subgroups. The aim of this study is to validate this molecular classification and test its prognostic relevance. Methods: Three hundred two patients were included in this study. Molecular analysis was performed for five CIMP-related promoters (CRABP1, MLH1, p16INK4a, CACNA1G, NEUROG1), MGMT, MSI, KRAS, and BRAF. Methylation in at least 4 promoters or in one to three promoters was considered CIMP-high and CIMP-low (CIMP-H/L), respectively. Results: CIMP-H, CIMP-L, and CIMP-negative were found in 7.1, 43, and 49.9% cases, respectively. One hundred twenty-three tumors (41%) could not be classified into any one of the proposed molecular subgroups, including 107 CIMP-L, 14 CIMP-H, and two CIMP-negative cases. The 10 year survival rate for CIMP-high patients [22.6% (95%CI: 7-43)] was significantly lower than for CIMP-L or CIMP-negative (p = 0.0295). Only the combined analysis of BRAF and CIMP (negative versus L/H) led to distinct prognostic subgroups. Conclusion: Although CIMP status has an effect on outcome, our results underline the need for standardized definitions of low- and high-level CIMP, which clearly hinders an effective prognostic and molecular classification of colorectal cancer.

Contemporary consensus proposal on criteria and classification of eosinophilic disorders and related syndromes

Eosinophilia is an important indicator of various neoplastic and nonneoplastic conditions. Depending on the underlying disease and mechanisms, eosinophil infiltration can lead to organ dysfunction, clinical symptoms, or both. During the past 2 decades, several different classifications of eosinophilic disorders and related syndromes have been proposed in various fields of medicine. Although criteria and definitions are, in part, overlapping, no global consensus has been presented to date. The Year 2011 Working Conference on Eosinophil Disorders and Syndromes was organized to update and refine the criteria and definitions for eosinophilic disorders and to merge prior classifications in a contemporary multidisciplinary schema. A panel of experts from the fields of immunology, allergy, hematology, and pathology contributed to this project. The expert group agreed on unifying terminologies and criteria and a classification that delineates various forms of hypereosinophilia, including primary and secondary variants based on specific hematologic and immunologic conditions, and various forms of the hypereosinophilic syndrome. For patients in whom no underlying disease or hypereosinophilic syndrome is found, the term hypereosinophilia of undetermined significance is introduced. The proposed novel criteria, definitions, and terminologies should assist in daily practice, as well as in the preparation and conduct of clinical trials.