Raul Santos-Rodriguez

An End-to-End Machine Learning System for Harmonic Analysis of Music

We present a new system for the harmonic analysis of popular musical audio. It is focused on chord estimation, although the proposed system additionally estimates the key sequence and bass notes. It is distinct from competing approaches in two main ways. First, it makes use of a new improved chromagram representation of audio that takes the human perception of loudness into account. Furthermore, it is the first system for joint estimation of chords, keys, and bass notes that is fully based on machine learning, requiring no expert knowledge to tune the parameters. This means that it will benefit from future increases in available annotated audio files, broadening its applicability to a wider range of genres. In all of three evaluation scenarios, including a new one that allows evaluation on audio for which no complete ground truth annotation is available, the proposed system is shown to be faster, more memory efficient, and more accurate than the state-of-the-art.

Automatic Chord Estimation from Audio: A Review of the State of the Art

In this overview article, we review research on the task of Automatic Chord Estimation (ACE). The major contributions from the last 14 years of research are summarized, with detailed discussions of the following topics: feature extraction, modeling strategies, model training and datasets, and evaluation strategies. Results from the annual benchmarking evaluation Music Information Retrieval Evaluation eXchange (MIREX) are also discussed as well as developments in software implementations and the impact of ACE within MIR. We conclude with possible directions for future research.

Spectral Clustering and Feature Selection for Microarray Data

Microarray datasets comprise a large number of gene expression values and a relatively small number of samples. Feature selection algorithms are very useful in these situations in order to find a compact subset of informative features. We propose a redundancy control method for algorithms in the recently proposed SPEC family of spectral-based feature selection algorithms. This method is applied to find relevant genes in order to cluster samples corresponding to three kinds of cancer: lung, breast and colon.

Cost-Sensitive Classification Based on Bregman Divergences for Medical Diagnosis

Medical applications, such as medical diagnosis, can be understood as classification problems. While usual approaches try to minimize the number of errors, medical scenarios often require classifiers that face up with different types of costs. This paper analyzes the application of a particular class of Bregman divergences to design cost sensitive classifiers for medical applications. It has been shown that these divergence measures can be used to estimate posterior probabilities with maximal accuracy for the probability values that are close to the decision boundaries. Experimental results on various medical datasets support the efficacy of our method.

Subjectively Interesting Component Analysis: Data Projections that Contrast with Prior Expectations

Methods that find insightful low-dimensional projections are essential to effectively explore high-dimensional data. Principal Component Analysis is used pervasively to find low-dimensional projections, not only because it is straightforward to use, but it is also often effective, because the variance in data is often dominated by relevant structure. However, even if the projections highlight real structure in the data, not all structure is interesting to every user. If a user is already aware of, or not interested in the dominant structure, Principal Component Analysis is less effective for finding interesting components. We introduce a new method called Subjectively Interesting Component Analysis (SICA), designed to find data projections that are subjectively interesting, i.e, projections that truly surprise the end-user. It is rooted in information theory and employs an explicit model of a users prior expectations about the data. The corresponding optimization problem is a simple eigenvalue problem, and the result is a trade-off between explained variance and novelty. We present five case studies on synthetic data, images, time-series, and spatial data, to illustrate how SICA enables users to find (subjectively) interesting projections.

Spatial/spectral information trade-off in hyperspectral images

This paper shows an empirical analysis of the trade-off between the spectral and the spatial information content of hyperspectral images. The objective of this study is to provide some insights into how changes and variations of both resolutions may affect the information content of the resulting image. This is useful for different stages of hyperspectral image processing: from acquisition to final applications. We propose two alternative approaches to measure the information content of a hyperspectral image: first, a second order approximation where the data distribution is supposed to be Gaussian, and secondly a higher order approximation where no assumption about the data distribution is made.

Cost-Sensitive Feature Selection Based on the Set Covering Machine

This paper describes how to make use of the cost information related to the extraction of each feature in a feature selection algorithm. For instance, in medical diagnosis, the different tests a patient might take during the diagnosis process can have different associated costs. The main idea is to change the feature selection framework in order to get low-cost subsets of informative features. This work proposes a way to introduce this information in a well-known machine learning algorithm, the Set Covering Machine.

SPHERE in a Box: Practical and Scalable EurValve Activity Monitoring Smart Home Kit

Non-invasive, environmental monitoring is being successfully utilised to improve health care outcomes for patients while allowing them to more safely and comfortably live in their homes instead of health care facilities. This promises to reduce costs and ease the health care burden for many countries globally. However, these systems are still in early stages of research and only highly skilled researchers and engineers can successfully deploy them. The difficulty in deploying these systems prevents their mass use and potential cost savings motivating research interest in smart homes in a box (SHiB). In this paper we present the EurValve Activity Monitoring Kit, a minimalist activity monitoring system that can be deployed in a home by the patient and still obtain valuable lifestyle and activity level information for medical clinicians. We describe the design of the system and how it is being used in the H2020 EurValve Project. The initial results show that the system is easily deployed and yet still effective for non-invasive sensing for activity classification and localisation.

Consistency of Losses for Learning from Weak Labels

In this paper we analyze the consistency of loss functions for learning from weakly labelled data, and its relation to properness. We show that the consistency of a given loss depends on the mixing matrix, which is the transition matrix relating the weak labels and the true class. A linear transformation can be used to convert a conventional classification-calibrated (CC) loss into a weak CC loss. By comparing the maximal dimension of the set of mixing matrices that are admissible for a given CC loss with that for proper losses, we show that classification calibration is a much less restrictive condition than properness. Moreover, we show that while the transformation of conventional proper losses into a weak proper losses does not preserve convexity in general, conventional convex CC losses can be easily transformed into weak and convex CC losses. Our analysis provides a general procedure to construct convex CC losses, and to identify the set of mixing matrices admissible for a given transformation. Several examples are provided to illustrate our approach.

A machine learning approach to intensive care discharge.

Objective The primary objective is to work towards a clinical decision support tool that can improve discharge practice on the intensive care unit. Design We used two datasets of routinely collected patient data to test and improve upon a set of previously proposed discharge criteria. Setting Bristol Royal Infirmary general intensive care unit (GICU). Patients Two cohorts derived from historical datasets: 1933 intensive care patients from GICU in Bristol, and 10658 from MIMIC-III (a publicly available intensive care dataset). Interventions None. Primary outcome measure None Results In both cohorts few successfully discharged patients met the of all the discharge criteria. Both a random forest and a logistic classifier, trained on MIMIC and cross validated on GICU, demonstrated improved performance over the original criteria and generalised well between the cohorts. The classifiers showed good agreement on which features were most predictive of readiness-for-discharge, and these were generally consistent with clinical experience. By weighting the NLD criteria according to feature importance from the logistic model we showed improved performance over the original NLD criteria, while retaining good interpretability. Conclusions Our findings constitute a proof of concept for a decision support tool to run alongside a clinical information system, and streamline the process of discharge from the ICU. Strengths and Limitations of this study This study applies machine learning techniques to the problem of classifying patients that are ready for discharge from intensive care. Two cohorts of historical data are used, allowing cross-validation and a comparison of results between healthcare contexts. Our approach represents the first step towards a decision support tool that would help clinicians identify dischargeable patients as early as possible.Abstract Objective The primary objective is to develop an automated method for detecting patients that are ready for discharge from intensive care. Design We used two datasets of routinely collected patient data to test and improve upon a set of previously proposed discharge criteria. Setting Bristol Royal Infirmary general intensive care unit (GICU). Patients Two cohorts derived from historical datasets: 1870 intensive care patients from GICU in Bristol, and 7592 from MIMIC-III (a publicly available intensive care dataset). Results In both cohorts few successfully discharged patients met all of the discharge criteria. Both a random forest and a logistic classifier, trained using multiple-source cross-validation, demonstrated improved performance over the original criteria and generalised well between the cohorts. The classifiers showed good agreement on which features were most predictive of readiness-for-discharge, and these were generally consistent with clinical experience. By weighting the discharge criteria according to feature importance from the logistic model we showed improved performance over the original criteria, while retaining good interpretability. Conclusions Our findings indicate the feasibility of the proposed approach to ready-for-discharge classification, which could complement other risk models of specific adverse outcomes in a future decision support system. Avenues for improvement to produce a clinically useful tool are identified. Strengths and Limitations of this study Training data from multiple source domains is leveraged to produce general classifiers. The restrictive feature representation tested could be expanded to better exploit the richness of available data and boost performance. Our approach has the potential to streamline the discharge process in cases where patient physiology makes them clear candidates for a de-escalation of care. High-risk patients would require additional levels of decision support to facilitate complex discharge planning.