Using unsupervised analysis to constrain generalization bounds for support vector classifiers

Meta-analysis informed machine learning: Supporting cytokine storm detection during CAR-T cell Therapy

Cytokine release syndrome (CRS), also known as cytokine storm, is one of the most consequential adverse effects of chimeric antigen receptor therapies that have shown otherwise promising results in cancer treatment. When emerging, CRS could be identified by the analysis of specific cytokine and chemokine profiles that tend to exhibit similarities across patients. In this paper, we exploit these similarities using machine learning algorithms and set out to pioneer a meta-review informed method for the identification of CRS based on specific cytokine peak concentrations and evidence from previous clinical studies. To this end we also address a widespread challenge of the applicability of machine learning in general: reduced training data availability. We do so by augmenting available (but often insufficient) patient cytokine concentrations with statistical knowledge extracted from domain literature. We argue that such methods could support clinicians in analyzing suspect cytokine profiles by matching them against the said CRS knowledge from past clinical studies, with the ultimate aim of swift CRS diagnosis. We evaluate our proposed methods under several design choices, achieving performance of more than 90% in terms of CRS identification accuracy, and showing that many of our choices outperform a purely data-driven alternative. During evaluation with real-world CRS clinical data, we emphasize the potential of our proposed method of producing interpretable results, in addition to being effective in identifying the onset of cytokine storm.

Opportunities and Challenges for Machine Learning in Rare Diseases

Rare diseases (RDs) are complicated health conditions that are difficult to be managed at several levels. The scarcity of available data chiefly determines an intricate scenario even for experts and specialized clinicians, which in turn leads to the so called “diagnostic odyssey” for the patient. This situation calls for innovative solutions to support the decision process via quantitative and automated tools. Machine learning brings to the stage a wealth of powerful inference methods; however, matching the health conditions with advanced statistical techniques raises methodological, technological, and even ethical issues. In this contribution, we critically point to the specificities of the dialog of rare diseases with machine learning techniques concentrating on the key steps and challenges that may hamper or create actionable knowledge and value for the patient together with some on-field methodological suggestions and considerations.

Finding Principal Paths in Data Space

In this paper, we introduce the concept of principal paths in data space; we show that this is a well-characterized problem from the point of view of cognition, and that it can lead to salient insights in the analyzed data enabling topological/holistic descriptions. These paths, interestingly, can be interpreted as local principal curves, and in this paper, we suggest that they are analogous to what, in the statistical mechanics realm, are called minimum free-energy paths. Here, we move that concept from physics to data space and compute them in both the original and the kernel space. The algorithm is a regularized version of the well-known k -means clustering algorithm. The regularization parameter is derived via an in-sample model selection process based on the Bayesian evidence maximization. Interestingly, we show that this choice for the regularization parameter consistently leads to the same manifold even when changing the number of clusters. We apply the method to common data sets, dynamical systems, and, in particular, to molecular dynamics trajectories showing the generality, the usefulness of the approach and its superiority with respect to other related approaches.

Computational intelligence techniques for tactile sensing systems

Tactile sensing helps robots interact with humans and objects effectively in real environments. Piezoelectric polymer sensors provide the functional building blocks of the robotic electronic skin, mainly thanks to their flexibility and suitability for detecting dynamic contact events and for recognizing the touch modality. The paper focuses on the ability of tactile sensing systems to support the challenging recognition of certain qualities/modalities of touch. The research applies novel computational intelligence techniques and a tensor-based approach for the classification of touch modalities; its main results consist in providing a procedure to enhance system generalization ability and architecture for multi-class recognition applications. An experimental campaign involving 70 participants using three different modalities in touching the upper surface of the sensor array was conducted, and confirmed the validity of the approach.

Unsupervised learning of categorical data with competing models

This paper considers the unsupervised learning of high-dimensional binary feature vectors representing categorical information. A cognitively inspired framework, referred to as modeling fields theory (MFT), is utilized as the basic methodology. A new MFT-based algorithm, referred to as accelerated maximum a posteriori (MAP), is proposed. Accelerated MAP allows simultaneous learning and selection of the number of models. The key feature of accelerated MAP is a steady increase of the regularization penalty resulting in competition among models. The differences between this approach and other mixture learning and model selection methodologies are described. The operation of this algorithm and its parameter selection are discussed. Numerical experiments aimed at finding performance limits are conducted. The performance with real-world data is tested by applying the algorithm to a text categorization problem and to the clustering Congressional voting data.

Regularization path for ν-support vector classification

The v-support vector classification (v-SVC) proposed by Schölkopf has the advantage of using a regularization parameter v for controlling the number of support vectors and margin errors. However, compared to C-SVC, its formulation is more complicated, and to date there are no effective methods for computing its regularization path. In this paper, we propose a new regularization path algorithm, which is designed on the basis of a modified formulation of v-SVC and traces the solution path with respect to the parameter v. Through theoretical analysis and confirmatory experiments, we show that our algorithm can avoid the infeasible updating path under several assumptions (i.e., Assumptions 1 and 2), and fit the entire solution path in a finite number of steps. When the regularization path of v-SVC is available, a novel approach proposed by Yang and Ong can be applied to obtain the global optimal solution of common validation functions for v-SVC, and the computation for the whole process is minimal. Numerical experiments show that it is more efficient than various kinds of grid search methods for selecting the optimal regularization parameter v.

Determination of global minima of some common validation functions in support vector machine

Tuning of the regularization parameter C is a well-known process in the implementation of a support vector machine (SVM) classifier. Such a tuning process uses an appropriate validation function whose value, evaluated over a validation set, has to be optimized for the determination of the optimal C. Unfortunately, most common validation functions are not smooth functions of C. This brief presents a method for obtaining the global optimal solution of these non-smooth validation functions. The method is guaranteed to find the global optimum and relies on the regularization solution path of SVM over a range of C values. When the solution path is available, the computation needed is minimal.

Mixing linear SVMs for nonlinear classification

In this paper, we address the problem of combining linear support vector machines (SVMs) for classification of large-scale nonlinear datasets. The motivation is to exploit both the efficiency of linear SVMs (LSVMs) in learning and prediction and the power of nonlinear SVMs in classification. To this end, we develop a LSVM mixture model that exploits a divide-and-conquer strategy by partitioning the feature space into subregions of linearly separable datapoints and learning a LSVM for each of these regions. We do this implicitly by deriving a generative model over the joint data and label distributions. Consequently, we can impose priors on the mixing coefficients and do implicit model selection in a top-down manner during the parameter estimation process. This guarantees the sparsity of the learned model. Experimental results show that the proposed method can achieve the efficiency of LSVMs in the prediction phase while still providing a classification performance comparable to nonlinear SVMs.

Behavior-constrained support vector machines for fMRI data analysis

Statistical learning methods are emerging as a valuable tool for decoding information from neural imaging data. The noisy signal and the limited number of training patterns that are typically recorded from functional brain imaging experiments pose a challenge for the application of statistical learning methods in the analysis of brain data. To overcome this difficulty, we propose using prior knowledge based on the behavioral performance of human observers to enhance the training of support vector machines (SVMs). We collect behavioral responses from human observers performing a categorization task during functional magnetic resonance imaging scanning. We use the psychometric function generated based on the observers behavioral choices as a distance constraint for training an SVM. We call this method behavior-constrained SVM (BCSVM). Our findings confirm that BCSVM outperforms SVM consistently.