Results 1 
7 of
7
Making machine learning models interpretable
 In Proc. European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning
, 2012
"... Abstract. Data of different levels of complexity and of ever growing diversity of characteristics are the raw materials that machine learning practitioners try to model using their wide palette of methods and tools. The obtained models are meant to be a synthetic representation of the available, obs ..."
Abstract

Cited by 10 (2 self)
 Add to MetaCart
(Show Context)
Abstract. Data of different levels of complexity and of ever growing diversity of characteristics are the raw materials that machine learning practitioners try to model using their wide palette of methods and tools. The obtained models are meant to be a synthetic representation of the available, observed data that captures some of their intrinsic regularities or patterns. Therefore, the use of machine learning techniques for data analysis can be understood as a problem of pattern recognition or, more informally, of knowledge discovery and data mining. There exists a gap, though, between data modeling and knowledge extraction. Models, depending on the machine learning techniques employed, can be described in diverse ways but, in order to consider that some knowledge has been achieved from their description, we must take into account the human cognitive factor that any knowledge extraction process entails. These models as such can be rendered powerless unless they can be interpreted, andthe process of human interpretation follows rules that go well beyond technical prowess. For this reason, interpretability is a paramount quality that machine learning methods should aim to achieve if they are to be applied in practice. This paper is a brief introduction to the special session on interpretable models in machine learning, organized as part of the 20 th European Symposium on Artificial Neural Networks, Computational Intelligence and Machine Learning. It includes a discussion on the several works accepted for the session, with an overview of the context of wider research on interpretability of machine learning models. 1
Cartogram representation of the batchSOM magnification factor
 IN EUROPEAN SYMPOSIUM ON ARTIFICIAL NEURAL NETWORKS (ESANN
, 2012
"... Model interpretability is a problem of knowledge extraction from the patterns found in raw data. One key source of knowledge is information visualization, which can help us to gain insights into a problem through graphical representations and metaphors. Nonlinear dimensionality reduction techniqu ..."
Abstract

Cited by 4 (4 self)
 Add to MetaCart
(Show Context)
Model interpretability is a problem of knowledge extraction from the patterns found in raw data. One key source of knowledge is information visualization, which can help us to gain insights into a problem through graphical representations and metaphors. Nonlinear dimensionality reduction techniques can provide flexible visual insight, but the locally varying representation distortion they produce makes interpretation far from intuitive. In this paper, we define a cartogram method, based on techniques of geographic representation, that allows reintroducing this distortion, measured as a magnification factor, in the visual maps of the batchSOM model. It does so while preserving the topological continuity of the representation.
Research directions in interpretable machine learning models
 In European Symposium on Artificial Neuronal Networks, Computational Intelligence and Machiene Learning
, 2013
"... Abstract. The theoretical novelty of many machine learning methods leading to high performing algorithms has been substantial. However, the blackbox nature of much of this body of work has meant that the models are difficult to interpret, with the consequence that the significant developments in ma ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
Abstract. The theoretical novelty of many machine learning methods leading to high performing algorithms has been substantial. However, the blackbox nature of much of this body of work has meant that the models are difficult to interpret, with the consequence that the significant developments in machine learning theory are not matched by their practical impact. This tutorial stresses the need for interpretation and outlines the current status and future directions of interpretability in machine learning models. 1 Why interpretation and visualization in machine learning? The above question directly corresponds in many applications to asking – why should machine learning methods be useful in practice? While there are many publications in this huge and significant field of learning, realworld applications are much fewer, especially in safetycritical domains. What are the reasons for this? How can flexible nonlinear models be interpreted? Alternatively, given that there are different ways of articulating a flexible regression or classification model, can machine learning models be designed so that they are directly interpretable by construction? Is interpretation in
Visualization and Interpretability in Probabilistic Dimensionality Reduction Models
, 2014
"... Over the last few decades, data analysis has swiftly evolved from being a task addressed mainly within the remit of multivariate statistics, to an endevour in which data heterogeneity, complexity and even sheer size, driven by computational advances, call for alternative strategies, such as those ..."
Abstract
 Add to MetaCart
(Show Context)
Over the last few decades, data analysis has swiftly evolved from being a task addressed mainly within the remit of multivariate statistics, to an endevour in which data heterogeneity, complexity and even sheer size, driven by computational advances, call for alternative strategies, such as those provided by pattern recognition and machine learning. Any data analysis process aims to extract new knowledge from data. Knowledge extraction is not a trivial task and it is not limited to the generation of data models or the recognition of patterns. The use of machine learning techniques for multivariate data analysis should in fact aim to achieve a dual target: interpretability and good performance. At best, both aspects of this target should not conflict with each other. This gap between data modelling and knowledge extraction must be acknowledged, in the sense that we can only extract knowledge from models through a process of interpretation. Exploratory information visualization is becoming a very promising tool for in
Local metric and graph based distance for Probabilistic Dimensionality Reduction
"... Probabilistic Dimensionality Reduction methods can provide a flexible data representation and a more faithful model of the observed Multivariate Datasets. This target is too often reached at the expense of model interpretability, which has an impact in the model visualization results. In many pract ..."
Abstract
 Add to MetaCart
(Show Context)
Probabilistic Dimensionality Reduction methods can provide a flexible data representation and a more faithful model of the observed Multivariate Datasets. This target is too often reached at the expense of model interpretability, which has an impact in the model visualization results. In many practical applications, an optimum performance could be less relevant than the achievement of interpretability: this is often the case in areas such as Medicine, Biology, Astronomy, Finance and Engineering (to name just a few). In this context, the task of data visualization is central to data exploration [1]. In manifold learning, when a highdimensional space is mapped onto a lowerdimensional one, the obtained embedded manifold is subject to some local geometrical distortion induced by the nonlinear mapping (manifold compression, stretching, gluing and tearing). This kind of distortion can often lead to misinterpretations of the data set itself. But, given that it is almost impossible to completely avoid geometrical distortions while reducing dimensionality, it is important to give relevance to another aspect of the problem: how to interpret the geometry and the local metric of the model in order to explore the data in a more faithful way. We consider here an explicit way to compute local metrics in generative models who perform probabilistic dimensionality reduction. The obtained metric tensor is here used to compute geodesic distances over the latent space using a graphbased dicretisation of the latent space itself. This way, the computed distances better reflects the underlying structure of the dataset. −2 −1
de Catalunya
"... We investigate the geometrical structure of probabilistic generative dimensionality reduction models using the tools of Riemannian geometry. We explicitly define a distribution over the natural metric given by the models. We provide the necessary algorithms to compute expected metric tensors where ..."
Abstract
 Add to MetaCart
We investigate the geometrical structure of probabilistic generative dimensionality reduction models using the tools of Riemannian geometry. We explicitly define a distribution over the natural metric given by the models. We provide the necessary algorithms to compute expected metric tensors where the distribution over mappings is given by a Gaussian process. We treat the corresponding latent variable model as a Riemannian manifold and we use the expectation of the metric under the Gaussian process prior to define interpolating paths and measure distance between latent points. We show how distances that respect the expected metric lead to more appropriate generation of new data. 1
Metrics for probabilistic geometry
, 2014
"... We investigate the geometrical structure of probabilistic generative dimensionality reduction models using the tools of Riemannian geometry. We explicitly define a distribution over the natural metric given by the models. We provide the necessary algorithms to compute expected metric tensors where t ..."
Abstract
 Add to MetaCart
We investigate the geometrical structure of probabilistic generative dimensionality reduction models using the tools of Riemannian geometry. We explicitly define a distribution over the natural metric given by the models. We provide the necessary algorithms to compute expected metric tensors where the distribution over mappings is given by a Gaussian process. We treat the corresponding latent variable model as a Riemannian manifold and we use the expectation of the metric under the Gaussian process prior to define interpolating paths and measure distance between latent points. We show how distances that respect the expected metric lead to more appropriate generation of new data.