Aquatic weed automatic classification using machine learning techniques

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Efficient supervised optimum-path forest classification for large datasets

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Improving the accuracy of the optimum-path forest supervised classifier for large datasets

In this work, a new approach for supervised pattern recognition is presented which improves the learning algorithm of the Optimum-Path Forest classifier (OPF), centered on detection and elimination of outliers in the training set. Identification of outliers is based on a penalty computed for each sample in the training set from the corresponding number of imputable false positive and false negative classification of samples. This approach enhances the accuracy of OPF while still gaining in classification time, at the expense of a slight increase in training time. © 2010 Springer-Verlag.

Precipitates segmentation from scanning electron microscope images through machine learning techniques

The presence of precipitates in metallic materials affects its durability, resistance and mechanical properties. Hence, its automatic identification by image processing and machine learning techniques may lead to reliable and efficient assessments on the materials. In this paper, we introduce four widely used supervised pattern recognition techniques to accomplish metallic precipitates segmentation in scanning electron microscope images from dissimilar welding on a Hastelloy C-276 alloy: Support Vector Machines, Optimum-Path Forest, Self Organizing Maps and a Bayesian classifier. Experimental results demonstrated that all classifiers achieved similar recognition rates with good results validated by an expert in metallographic image analysis. © 2011 Springer-Verlag Berlin Heidelberg.

Preliminary diagnosis of ophtalmological diseases through machine learning techniques

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

Word sense disambiguation via high order of learning in complex networks

Complex networks have been employed to model many real systems and as a modeling tool in a myriad of applications. In this paper, we use the framework of complex networks to the problem of supervised classification in the word disambiguation task, which consists in deriving a function from the supervised (or labeled) training data of ambiguous words. Traditional supervised data classification takes into account only topological or physical features of the input data. On the other hand, the human (animal) brain performs both low- and high-level orders of learning and it has facility to identify patterns according to the semantic meaning of the input data. In this paper, we apply a hybrid technique which encompasses both types of learning in the field of word sense disambiguation and show that the high-level order of learning can really improve the accuracy rate of the model. This evidence serves to demonstrate that the internal structures formed by the words do present patterns that, generally, cannot be correctly unveiled by only traditional techniques. Finally, we exhibit the behavior of the model for different weights of the low- and high-level classifiers by plotting decision boundaries. This study helps one to better understand the effectiveness of the model. Copyright (C) EPLA, 2012

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.

Learning by the Dendritic Prediction of Somatic Spiking

Recent modeling of spike-timing-dependent plasticity indicates that plasticity involves as a third factor a local dendritic potential, besides pre- and postsynaptic firing times. We present a simple compartmental neuron model together with a non-Hebbian, biologically plausible learning rule for dendritic synapses where plasticity is modulated by these three factors. In functional terms, the rule seeks to minimize discrepancies between somatic firings and a local dendritic potential. Such prediction errors can arise in our model from stochastic fluctuations as well as from synaptic input, which directly targets the soma. Depending on the nature of this direct input, our plasticity rule subserves supervised or unsupervised learning. When a reward signal modulates the learning rate, reinforcement learning results. Hence a single plasticity rule supports diverse learning paradigms.

Livelihood and Community Development: LforS learning and simulation game

Up to 15 people can participate in the game, which is supervised by a moderator. Households consisting of 1-5 people discuss options for diversification of household strategies. Aim of the game: By devising appropriate strategies, households seek to stand up to various types of events while improving their economic and social situation and, at the same time, taking account of ecological conditions. The annual General Community Meeting (GCM) provides an opportunity for households to create a general set-up at the local level that is more or less favourable to the strategies they are pursuing. The development of a community investment strategy, to be implemented by the GCM, and successful coordination between households will allow players to optimise their investments at the household level. The household who owns the most assets at the end of the game wins. Players participate very actively, as the game stimulates lively and interesting discussions. They find themselves confronted with different types of decision-making related to the reality of their daily lives. They explore different ways to model their own household strategies and discuss risks and opportunities. Reflections on the course of the game continually refer to the real-life situations of the participants.

Regularization for sparsity in statistical analysis and machine learning

Pragmatism is the leading motivation of regularization. We can understand regularization as a modification of the maximum-likelihood estimator so that a reasonable answer could be given in an unstable or ill-posed situation. To mention some typical examples, this happens when fitting parametric or non-parametric models with more parameters than data or when estimating large covariance matrices. Regularization is usually used, in addition, to improve the bias-variance tradeoff of an estimation. Then, the definition of regularization is quite general, and, although the introduction of a penalty is probably the most popular type, it is just one out of multiple forms of regularization. In this dissertation, we focus on the applications of regularization for obtaining sparse or parsimonious representations, where only a subset of the inputs is used. A particular form of regularization, L1-regularization, plays a key role for reaching sparsity. Most of the contributions presented here revolve around L1-regularization, although other forms of regularization are explored (also pursuing sparsity in some sense). In addition to present a compact review of L1-regularization and its applications in statistical and machine learning, we devise methodology for regression, supervised classification and structure induction of graphical models. Within the regression paradigm, we focus on kernel smoothing learning, proposing techniques for kernel design that are suitable for high dimensional settings and sparse regression functions. We also present an application of regularized regression techniques for modeling the response of biological neurons. Supervised classification advances deal, on the one hand, with the application of regularization for obtaining a na¨ıve Bayes classifier and, on the other hand, with a novel algorithm for brain-computer interface design that uses group regularization in an efficient manner. Finally, we present a heuristic for inducing structures of Gaussian Bayesian networks using L1-regularization as a filter. El pragmatismo es la principal motivación de la regularización. Podemos entender la regularización como una modificación del estimador de máxima verosimilitud, de tal manera que se pueda dar una respuesta cuando la configuración del problema es inestable. A modo de ejemplo, podemos mencionar el ajuste de modelos paramétricos o no paramétricos cuando hay más parámetros que casos en el conjunto de datos, o la estimación de grandes matrices de covarianzas. Se suele recurrir a la regularización, además, para mejorar el compromiso sesgo-varianza en una estimación. Por tanto, la definición de regularización es muy general y, aunque la introducción de una función de penalización es probablemente el método más popular, éste es sólo uno de entre varias posibilidades. En esta tesis se ha trabajado en aplicaciones de regularización para obtener representaciones dispersas, donde sólo se usa un subconjunto de las entradas. En particular, la regularización L1 juega un papel clave en la búsqueda de dicha dispersión. La mayor parte de las contribuciones presentadas en la tesis giran alrededor de la regularización L1, aunque también se exploran otras formas de regularización (que igualmente persiguen un modelo disperso). Además de presentar una revisión de la regularización L1 y sus aplicaciones en estadística y aprendizaje de máquina, se ha desarrollado metodología para regresión, clasificación supervisada y aprendizaje de estructura en modelos gráficos. Dentro de la regresión, se ha trabajado principalmente en métodos de regresión local, proponiendo técnicas de diseño del kernel que sean adecuadas a configuraciones de alta dimensionalidad y funciones de regresión dispersas. También se presenta una aplicación de las técnicas de regresión regularizada para modelar la respuesta de neuronas reales. Los avances en clasificación supervisada tratan, por una parte, con el uso de regularización para obtener un clasificador naive Bayes y, por otra parte, con el desarrollo de un algoritmo que usa regularización por grupos de una manera eficiente y que se ha aplicado al diseño de interfaces cerebromáquina. Finalmente, se presenta una heurística para inducir la estructura de redes Bayesianas Gaussianas usando regularización L1 a modo de filtro.

Fusion of probabilistic knowledge-based classification rules and learning automata for automatic recognition of digital images

In this paper, the fusion of probabilistic knowledge-based classification rules and learning automata theory is proposed and as a result we present a set of probabilistic classification rules with self-learning capability. The probabilities of the classification rules change dynamically guided by a supervised reinforcement process aimed at obtaining an optimum classification accuracy. This novel classifier is applied to the automatic recognition of digital images corresponding to visual landmarks for the autonomous navigation of an unmanned aerial vehicle (UAV) developed by the authors. The classification accuracy of the proposed classifier and its comparison with well-established pattern recognition methods is finally reported.

Machine Learning in Scientometrics

El aprendizaje automático y la cienciometría son las disciplinas científicas que se tratan en esta tesis. El aprendizaje automático trata sobre la construcción y el estudio de algoritmos que puedan aprender a partir de datos, mientras que la cienciometría se ocupa principalmente del análisis de la ciencia desde una perspectiva cuantitativa. Hoy en día, los avances en el aprendizaje automático proporcionan las herramientas matemáticas y estadísticas para trabajar correctamente con la gran cantidad de datos cienciométricos almacenados en bases de datos bibliográficas. En este contexto, el uso de nuevos métodos de aprendizaje automático en aplicaciones de cienciometría es el foco de atención de esta tesis doctoral. Esta tesis propone nuevas contribuciones en el aprendizaje automático que podrían arrojar luz sobre el área de la cienciometría. Estas contribuciones están divididas en tres partes: Varios modelos supervisados (in)sensibles al coste son aprendidos para predecir el éxito científico de los artículos y los investigadores. Los modelos sensibles al coste no están interesados en maximizar la precisión de clasificación, sino en la minimización del coste total esperado derivado de los errores ocasionados. En este contexto, los editores de revistas científicas podrían disponer de una herramienta capaz de predecir el número de citas de un artículo en el fututo antes de ser publicado, mientras que los comités de promoción podrían predecir el incremento anual del índice h de los investigadores en los primeros años. Estos modelos predictivos podrían allanar el camino hacia nuevos sistemas de evaluación. Varios modelos gráficos probabilísticos son aprendidos para explotar y descubrir nuevas relaciones entre el gran número de índices bibliométricos existentes. En este contexto, la comunidad científica podría medir cómo algunos índices influyen en otros en términos probabilísticos y realizar propagación de la evidencia e inferencia abductiva para responder a preguntas bibliométricas. Además, la comunidad científica podría descubrir qué índices bibliométricos tienen mayor poder predictivo. Este es un problema de regresión multi-respuesta en el que el papel de cada variable, predictiva o respuesta, es desconocido de antemano. Los índices resultantes podrían ser muy útiles para la predicción, es decir, cuando se conocen sus valores, el conocimiento de cualquier valor no proporciona información sobre la predicción de otros índices bibliométricos. Un estudio bibliométrico sobre la investigación española en informática ha sido realizado bajo la cultura de publicar o morir. Este estudio se basa en una metodología de análisis de clusters que caracteriza la actividad en la investigación en términos de productividad, visibilidad, calidad, prestigio y colaboración internacional. Este estudio también analiza los efectos de la colaboración en la productividad y la visibilidad bajo diferentes circunstancias. ABSTRACT Machine learning and scientometrics are the scientific disciplines which are covered in this dissertation. Machine learning deals with the construction and study of algorithms that can learn from data, whereas scientometrics is mainly concerned with the analysis of science from a quantitative perspective. Nowadays, advances in machine learning provide the mathematical and statistical tools for properly working with the vast amount of scientometrics data stored in bibliographic databases. In this context, the use of novel machine learning methods in scientometrics applications is the focus of attention of this dissertation. This dissertation proposes new machine learning contributions which would shed light on the scientometrics area. These contributions are divided in three parts: Several supervised cost-(in)sensitive models are learned to predict the scientific success of articles and researchers. Cost-sensitive models are not interested in maximizing classification accuracy, but in minimizing the expected total cost of the error derived from mistakes in the classification process. In this context, publishers of scientific journals could have a tool capable of predicting the citation count of an article in the future before it is published, whereas promotion committees could predict the annual increase of the h-index of researchers within the first few years. These predictive models would pave the way for new assessment systems. Several probabilistic graphical models are learned to exploit and discover new relationships among the vast number of existing bibliometric indices. In this context, scientific community could measure how some indices influence others in probabilistic terms and perform evidence propagation and abduction inference for answering bibliometric questions. Also, scientific community could uncover which bibliometric indices have a higher predictive power. This is a multi-output regression problem where the role of each variable, predictive or response, is unknown beforehand. The resulting indices could be very useful for prediction purposes, that is, when their index values are known, knowledge of any index value provides no information on the prediction of other bibliometric indices. A scientometric study of the Spanish computer science research is performed under the publish-or-perish culture. This study is based on a cluster analysis methodology which characterizes the research activity in terms of productivity, visibility, quality, prestige and international collaboration. This study also analyzes the effects of collaboration on productivity and visibility under different circumstances.

Learning higher-level features with convolutional restricted Boltzmann machines for sentiment analysis

In recent years, learning word vector representations has attracted much interest in Natural Language Processing. Word representations or embeddings learned using unsupervised methods help addressing the problem of traditional bag-of-word approaches which fail to capture contextual semantics. In this paper we go beyond the vector representations at the word level and propose a novel framework that learns higher-level feature representations of n-grams, phrases and sentences using a deep neural network built from stacked Convolutional Restricted Boltzmann Machines (CRBMs). These representations have been shown to map syntactically and semantically related n-grams to closeby locations in the hidden feature space. We have experimented to additionally incorporate these higher-level features into supervised classifier training for two sentiment analysis tasks: subjectivity classification and sentiment classification. Our results have demonstrated the success of our proposed framework with 4% improvement in accuracy observed for subjectivity classification and improved the results achieved for sentiment classification over models trained without our higher level features.

Parallel Semi-supervised EM-algorithm (MT-SSEM)

Report published in the Proceedings of the National Conference on "Education in the Information Society", Plovdiv, May, 2013

Learning Binary Code Representations for Effective and Efficient Image Retrieval

The size of online image datasets is constantly increasing. Considering an image dataset with millions of images, image retrieval becomes a seemingly intractable problem for exhaustive similarity search algorithms. Hashing methods, which encodes high-dimensional descriptors into compact binary strings, have become very popular because of their high efficiency in search and storage capacity. In the first part, we propose a multimodal retrieval method based on latent feature models. The procedure consists of a nonparametric Bayesian framework for learning underlying semantically meaningful abstract features in a multimodal dataset, a probabilistic retrieval model that allows cross-modal queries and an extension model for relevance feedback. In the second part, we focus on supervised hashing with kernels. We describe a flexible hashing procedure that treats binary codes and pairwise semantic similarity as latent and observed variables, respectively, in a probabilistic model based on Gaussian processes for binary classification. We present a scalable inference algorithm with the sparse pseudo-input Gaussian process (SPGP) model and distributed computing. In the last part, we define an incremental hashing strategy for dynamic databases where new images are added to the databases frequently. The method is based on a two-stage classification framework using binary and multi-class SVMs. The proposed method also enforces balance in binary codes by an imbalance penalty to obtain higher quality binary codes. We learn hash functions by an efficient algorithm where the NP-hard problem of finding optimal binary codes is solved via cyclic coordinate descent and SVMs are trained in a parallelized incremental manner. For modifications like adding images from an unseen class, we propose an incremental procedure for effective and efficient updates to the previous hash functions. Experiments on three large-scale image datasets demonstrate that the incremental strategy is capable of efficiently updating hash functions to the same retrieval performance as hashing from scratch.