|
|
Multi-model Approach For Effective Multimedia Exploration
Grošup, Tomáš ; Lokoč, Jakub (advisor) ; Somol, Petr (referee)
This work is focusing on exploration of multimedia collections. It describes the problematic of exploration and proposes new approaches to it, two based on the data structure M-Index and two utilizing multiple similarity models at once. Those approaches were compared using an extensive user study. Part of this work is also devoted to analysis of a new exploration system, design of its architecture, system implementation and its deployment. This exploration system was used in several applications, which are also shown and described in this thesis.
|
|
|
A Comparison of Adaptive Sampling and Interpolation of 2D BRDF Subspaces
Vávra, Radomír ; Filip, Jiří ; Somol, P.
This report comprises overview of interpolation and sampling methods of Bidirectional Reflectance Distribution Function (BRDF). We analyzed 2D BRDF subspaces of eleven materials. We compared performance of five interpolation methods, three different sampling patterns, and compared twelve adaptive sampling strategies. Finally, based on knowledge of entire data we estimated sub-optimal sampling patterns and as a reference compared them with other tested sampling approaches.
|
|
| |
|
| |
|
|
Fast Dependency-Aware Feature Selection in Very-High-Dimensional Pattern Recognition Problems
Somol, Petr ; Grim, Jiří
The paper addresses the problem of making dependency-aware feature selection feasible in pattern recognition problems of very high dimensionality. The idea of individually best ranking is generalized to evaluate the contextual quality of each feature in a series of randomly generated feature subsets. Each random subset is evaluated by a criterion function of arbitrary choice (permitting functions of high complexity). Eventually, the novel dependency-aware feature rank is computed, expressing the average benefit of including a feature into feature subsets. The method is efficient and generalizes well especially in very-high-dimensional problems, where traditional context-aware feature selection methods fail due to prohibitive computational complexity or to over-fitting. The method is shown well capable of over-performing the commonly applied individual ranking which ignores important contextual information contained in data.
|
|
|
Introduction to Feature Selection Toolbox 3 – The C++ Library for Subset Search, Data Modeling and Classification
Somol, Petr ; Vácha, Pavel ; Mikeš, Stanislav ; Hora, Jan ; Pudil, Pavel ; Žid, Pavel
We introduce a new standalone widely applicable software library for feature selection (also known as attribute or variable selection), capable of reducing problem dimensionality to maximize the accuracy of data models, performance of automatic decision rules as well as to reduce data acquisition cost. The library can be exploited by users in research as well as in industry. Less experienced users can experiment with different provided methods and their application to real-life problems, experts can implement their own criteria or search schemes taking advantage of the toolbox framework. In this paper we first provide a concise survey of a variety of existing feature selection approaches. Then we focus on a selected group of methods of good general performance as well as on tools surpassing the limits of existing libraries. We build a feature selection framework around them and design an object-based generic software library. We describe the key design points and properties of the library.
|
|
|
Sequential Retreating Search Methods in Feature Selection
Somol, Petr ; Pudil, Pavel
Inspired by Floating Search, our new pair of methods, the Sequential Forward Retreating Search (SFRS) and Sequential Backward Retreating Search (SBRS) is exceptionally suitable for Wrapper based feature selection. (Conversely, it cannot be used with monotonic criteria.) Unlike most of other known sub-optimal search methods, both the SFRS and SBRS are parameter-free deterministic sequential procedures that incorporate in the optimization process both the search for the best subset and the determination of the best subset size. The subset yielded by either of the two new methods is to be expected closer to optimum than the best of all subsets yielded in one run of the Floating Search. Retreating Search time complexity is to be expected slightly worse but in the same order of magnitude as that of the Floating Search. In addition to introducing the new methods we provide a testing framework to evaluate them with respect to other existing tools.
|
|
|
Vyhodnocení stability jednotlivých metod i skupin metod výběru příznaků, který optimalizují kardinalitu podmnožiny příznaků
Somol, Petr ; Novovičová, Jana
Stability (robustness) of feature selection methods is a topic of recent interest yet often neglected importance with direct impact on the reliability of machine learning systems. We investigate the problem of evaluating the stability of feature selection processes yielding subsets of varying size. We introduce several novel feature selection stability measures and adjust some existing measures in a unifying framework that offers broad insight into the stability problem. We study in detail the properties of considered measures and demonstrate on various examples what information about the feature selection process can be gained. We also introduce an alternative approach to feature selection evaluation in form of measures that enable comparing the similarity of two feature selection processes. These measures enable comparing, e.g., the output of two feature selection methods or two runs of one method with different parameters. The information obtained using the considered stability and similarity measures is shown usable for assessing feature selection methods (or criteria) as such
|
|
|
Diagnostické vyhodnocování screeningových mamogramů pomocí lokálních texturních modelů
Grim, Jiří ; Somol, Petr
We propose statistically based preprocessing of screening mammograms with the aim to emphasize suspicious areas. We estimate the local statistical texture model of a single mammogram in the form of multivariate Gaussian mixture. The probability density is estimated from the data obtained by pixelwise scanning of the mammogram with the search window. In the second phase, we evaluate the estimated density at each position of the window and display the corresponding log-likelihood value as a gray level at the window center. Light gray levels correspond to the typical parts of the image and the dark values reflect unusual places. The resulting log-likelihood image exactly correlates with the structural details of the original mammogram, emphasizes locations of similar properties by contour lines and may provide additional information to facilitate diagnostic interpretation.
|
|
|
Má smysl vyvíjet nové metody výběru příznaků?
Somol, Petr ; Novovičová, Jana
One of hot topics discussed recently in relation to pattern recognition techniques is the question of actual performance of modern feature selection methods. Feature selection has been a highly active area of research in recent years due to its potential to improve both the performance and economy of automatic decision systems in various applicational fields, with medical diagnosis being among the most prominent. Feature selection may also improve the performance of classifiers learned from limited data, or contribute to model interpretability. The number of available methods and methodologies has grown rapidly while promising important improvements. Yet recently many authors put this development in question, claiming that simpler older tools show to be actually better than complex modern ones -- which, despite promises, are claimed to actually fail in real-world applications.
|
guest ::
RSS feed.