Quantitative analysis in a heuristic evaluation experiment

In usability context of interactive systems the heuristic evaluation method is widespread. In most applications the results tend to be qualitative, describing such aspects that require some improvement for the benefit of usability. However, these qualitative results do not allow us to determine how usable it is or how it becomes an interactive system. Hence the need for quantitative results may also be very necessary in order to determine the effort that would be needed to get a sufficiently usable system.This article describes, following the idea of the UsabAIPO Project, a new experiment to obtain quantitative results after a heuristic evaluation. This new experimentation has required some variation on the original idea, working with a set of different heuristic categories, while considering the use of the score depending on severity and frequency parameters.     

Expanding the landscape of biological computation with synthetic multicellular consortia

Computation is an intrinsic attribute of biological entities. All of them gather and process information and respond in predictable ways to an uncertain external environment. Are these computations similar to those performed by artificial systems? Can a living computer be constructed following standard engineering practices? Despite the similarities between molecular networks associated to information processing and the wiring diagrams used to represent electronic circuits, major differences arise. Such differences are specially relevant while engineering molecular circuits in order to build novel functionalities. Among others, wiring molecular components within a cell becomes a great challenge as soon as the complexity of the circuit becomes larger than simple gates. An alternative approach has been recently introduced based on a non-standard approach to cellular computation. By breaking some standard assumptions of engineering design, it allows the synthesis of multicellular engineered circuits able to perform complex functions and open a novel form of computation. Here we review previous studies dealing with both natural and synthetic forms of computation. We compare different systems spanning many spatial and temporal scales and outline a possible “space” of biological forms of computation. We suggest that a novel approach to build synthetic devices using multicellular consortia allows expanding this space in new directions.     

Trade‐Offs in Seed Dispersal Strategies Across Riparian Trees: The How Matters as Much as the When

Riparian Salicaceae are prolific producers of short‐lived seeds that require very restrictive hydro‐geomorphic conditions for establishment. It is generally assumed that if floods are able to create nursery sites timed with seed dispersal, recruitment will occur. Other spatial and temporal seed dispersal patterns besides the dispersal period have historically received little attention. However, seed dispersal patterns can be highly variable between regions, species and over the years. In this paper, we report the seed dispersal patterns of three dominant riparian Salicaceae trees in Europe: Populus alba, P. nigra and Salix alba to suggest possible trade‐offs between seed dispersal patterns, germinability, longevity and establishment. Seed rain of the three species was monitored in 33 glue‐coated traps for three months yearly from 2006 to 2008 in an 8‐km stretch of the Middle Ebro River (N Spain), which has a pluvio‐nival regime. P. alba dispersed seeds earlier during a shorter time period and with a fewer number of seed release pulses compared with P. nigra, and especially with S. alba. With overlapping seed dispersal periods, the two latter species occupy similar landform units but rarely compete with P. alba, usually at higher elevations, as shown in a previous study in the same study area. The three species had very high germinability immediately after release (>90%), but longevity in S. alba was eight times shorter than that of its two Populus counterparts. We suggest that S. alba has compensated its lower seed quality with a more spaced seed release in several pulses of similar magnitude. With similar seed dispersal patterns and germinability but a higher longevity, P. nigra had a much higher density of individuals than S. alba in the recruitment zones of the study area. We hope that our results may inform river managers about how to optimize river flows to promote sexual regeneration of these species. Copyright © 2015 John Wiley & Sons, Ltd.     

Online error correcting output codes

This article proposes a general extension of the error correcting output codes framework to the online learning scenario. As a result, the final classifier handles the addition of new classes independently of the base classifier used. In particular, this extension supports the use of both online example incremental and batch classifiers as base learners. The extension of the traditional problem independent codings one-versus-all and one-versus-one is introduced. Furthermore, two new codings are proposed, unbalanced online ECOC and a problem dependent online ECOC. This last online coding technique takes advantage of the problem data for minimizing the number of dichotomizers used in the ECOC framework while preserving a high accuracy. These techniques are validated on an online setting of 11 data sets from UCI database and applied to two real machine vision applications: traffic sign recognition and face recognition. As a result, the online ECOC techniques proposed provide a feasible and robust way for handling new classes using any base classifier.     

An ensemble-based method for linear feature extraction for two-class problems

In this paper we propose three variants of a linear feature extraction technique based on Adaboost for two-class classification problems. Unlike other feature extraction techniques, we do not make any assumptions about the distribution of the data. At each boosting step we select from a pool of linear projections the one that minimizes the weighted error. We propose three different variants of the feature extraction algorithm, depending on the way the pool of individual projections is constructed. Using nine real and two artificial data sets of different original dimensionality and sample size we compare the performance of the three proposed techniques with three classical techniques for linear feature extraction: Fisher linear discriminant analysis (FLD), Nonparametric discriminant analysis (NDA) and a recently proposed feature extraction method for heteroscedastic data based on the Chernoff criterion. Our results show that for data sets of relatively low-original dimensionality FLD appears to be both the most accurate and the most economical feature extraction method (giving just one-dimension in the case of two classes). The techniques based on Adaboost fare better than the classical techniques for data sets of large original dimensionality.

---

Geometry-Based Ensembles: Toward a Structural Characterization of the Classification Boundary

This article introduces a novel binary discriminative learning technique based on the approximation of the non-linear decision boundary by a piece-wise linear smooth additive model. The decision border is geometrically defined by means of the characterizing boundary points - points that belong to the optimal boundary under a certain notion of robustness. Based on these points, a set of locally robust linear classifiers is defined and assembled by means of a Tikhonov regularized optimization procedure in an additive model to create a final lambda-smooth decision rule. As a result, a very simple and robust classifier with a strong geometrical meaning and non-linear behavior is obtained. The simplicity of the method allows its extension to cope with some of nowadays machine learning challenges, such as online learning, large scale learning or parallelization, with linear computational complexity. We validate our approach on the UCI database. Finally, we apply our technique in online and large scale scenarios, and in six real life computer vision and pattern recognition problems: gender recognition, intravascular ultrasound tissue classification, speed traffic sign detection, Chagas' disease severity detection, clef classification and action recognition using a 3D accelerometer data. The results are promising and this paper opens a line of research that deserves further attention.     

Shared feature extraction for nearest neighbor face recognition.

In this paper, we propose a new supervised linear feature extraction technique for multiclass classification problems that is specially suited to the nearest neighbor classifier (NN). The problem of finding the optimal linear projection matrix is defined as a classification problem and the Adaboost algorithm is used to compute it in an iterative way. This strategy allows the introduction of a multitask learning (MTL) criterion in the method and results in a solution that makes no assumptions about the data distribution and that is specially appropriated to solve the small sample size problem. The performance of the method is illustrated by an application to the face recognition problem. The experiments show that the representation obtained following the multitask approach improves the classic feature extraction algorithms when using the NN classifier, especially when we have a few examples from each class.