This study tackles the problem of generating electrocardiograms (ECG) from photoplethysmograms (PPG) data collected by wearable devices such as smartwatches. Existing methods based on diffusion models successfully generate high-quality ECGs but suffer from slow inference since diffusion models iterate neural network processing to denoise noisy data. In this study, we propose a new solution, called CardioFlow, to this PPG-to-ECG translation problem; based on the recently proposed rectified flow framework, it can be interpreted as a one/few-step generative model. Furthermore, CardioFlow can computationally efficiently generate ECG-specific structures using masks highlighting PPG and ECG signal peaks. We conduct the experiments on two real-world biosignal datasets, WESAD and DALIA, and confirm that our method can generate high-quality ECGs faster than the existing diffusion-model-based methods.

Personalized interventions through dialog-based health guidance effectively encourage healthy behaviors for clients (i.e., individuals) at risk of lifestyle diseases. However, it is not obvious which dialog policies would enable the best understanding of the client’s characteristics and subsequent provision of tailored interventions. To meet this need, we propose a dialog policy learning method for both supporting professionals and directly motivating clients via chatbots. We model the dialog to identify tailored interventions, considering the heterogeneity among clients, by multitask reinforcement learning and train an agent in a model-based manner utilizing real-world data and reward shaping to ensure the safety of early agents’ statements. A simulation study utilizing real data shows that our approach achieves a success rate of 76% in identifying tailored interventions with few questions, outperforming standard methods, including large language models, which had a 42% success rate.

Shuffled regression is the problem of learning regression functions from shuffled data where the correspondence between the input features and target response is unknown. This paper proposes a probabilistic model for shuffled regression called Gaussian Process Shuffled Regression (GPSR). By introducing Gaussian processes as a prior of regression functions in function space via the kernel function, GPSR can express a wide variety of functions in a nonparametric manner while quantifying the uncertainty of the prediction. By adopting the Bayesian evidence maximization framework and a theoretical analysis of the connection between the marginal likelihood/predictive distribution of GPSR and that of standard Gaussian process regression (GPR), we derive an easy-to-implement inference algorithm for GPSR that iteratively applies GPR and updates the input-output correspondence. To reduce computation costs and obtain closed-form solutions for correspondence updates, we also develop a sparse approximate variant of GPSR using its weight space formulation, which can be seen as Bayesian shuffled linear regression with random Fourier features. Experiments on benchmark datasets confirm the effectiveness of our GPSR proposal.

Zero-shot learning is a promising approach to generalizing a model to categories unseen during training, and various methods have been proposed. However, they assume that class knowledge, the semantic information of the classes, is available as prior knowledge and thus fails to support domains whose class knowledge is unavailable. We propose a meta-learning method that allows us to apply the zero-shot learning method even if class knowledge is unavailable. We assume multiple zero-shot learning tasks where some classes are missing for each task, but they appear in other tasks. Our method simultaneously estimates the appropriate class knowledge and classification model using a meta-learning approach that extracts task features. We can use the trained models to perform zero-shot classification on unseen tasks without class knowledge. In experiments on datasets where true class knowledge is available, class knowledge is unavailable, and class knowledge is provided but imprecise, we show that the proposed method performs better than existing zero-shot learning methods.

Shuffled regression is the problem of learning regression models from shuffled data that consists of a set of input features and a set of target outputs where the correspondence between the input and output is unknown. This study proposes a new deep learning method for shuffled regression called Shuffled Deep Regression (SDR). We derive the sparse and stochastic variant of the Expectation-Maximization algorithm for SDR that iteratively updates discrete latent variables and the parameters of neural networks. The effectiveness of the proposal is confirmed by benchmark data experiments.

Dynamic mode decomposition (DMD) has been applied for analyzing various nonlinear dynamic systems. In this study, we clarify the relationship among the variants of DMD and nonnegative matrix factorization (NMF), which is also widely used for audio source separation and document clustering. We show that the decomposition form of DMD is a special case of complex NMF (CNMF), and both are equivalent when certain constraint conditions are added to CNMF. The equivalence of the objective functions of nonnegative DMD and CNMF under the aforementioned constraints is also proven. This theoretical analysis provides guidelines for an effective initial value setting of CNMF and for a new construction method that is positioned between CNMF and DMD.

This study proposes the latent neural phase model (LNPM), a deep generative model that can estimate the phase of coupled oscillators and parameters of the phase equation simultaneously from irregularly sampled noisy observation data. LNPM can handle a wide range of phase equations e.g., equations with time-dependent parameters. Inheriting the property of the classical phase model, we can use LNPM to quantify the degree of collective synchronization of rhythms. We develop a parameter estimation algorithm that minimizes the evidence lower bound (ELBO) by using the reparametrization trick. The effectiveness of LNPM is confirmed by challenges that include Huygens's clock problem.

The research field of stochastic matching has yielded many developments for various applications. In most stochastic matching problems, the probability distributions inherent in the nodes and edges are set a priori, and are not controllable. However, many matching services have options, which we call control variables, that affect the probability distributions and thus what constitutes an optimum matching. Although several methods for optimizing the values of the control variables have been developed, their optimization in consideration of the matching problem is still in its infancy. In this paper, we formulate an optimization problem for determining the values of the control variables so as to maximize the expected value of matching weights. Since this problem involves hard to evaluate objective values and is non-convex, we construct an approximation algorithm via a minimum-cost flow algorithm that can find 3-approximation solutions rapidly. Simulations on real data from a ride-hailing platform and a crowd-sourcing market show that the proposed method can find solutions with high profits of the service provider in practical time.

The importance of developing an application based on intervention technology that leads to a healthier life is widely recognized. A challenging part of realizing the application is the need for planning, i.e., considering a user's health goal (e.g., sleep at 10:00 p.m. to get enough sleep), providing intervention at the appropriate timing to help the user achieve the goal. The reinforcement learning (RL) approach is well suited to this type of problem since it is a methodology for planning; RL finds the optimal strategy as that which maximizes future expected profit. The purpose of this study is to clarify the effects of intervention based on RL to support healthy daily life. Therefore, we (i) collect real daily activity data from participants, (ii) generate a user model that imitates the user's response to system interventions, (iii) examine valuable goals and design them as rewards in RL and (iv) obtain optimal intervention strategies by RL via simulations given a user model and goals. We evaluate a generated user model and verify by simulations whether our method could successfully achieve the goal. In addition, we analyze the cases that demonstrated higher probability of achieving the goal and report the features.

Due to the difficulty of comprehensive data collection, created by factors such as privacy protection and sensor device limitations, we often need to analyze incomplete transition data where some information is missing from the ideal (complete) transition data. In this paper, we propose a new method that can estimate, in a unified manner, Markov chain parameters from incomplete transition data that consist of hidden transition data (data from which visited state information is partially hidden) and dropped transition data (data from which some state visits are dropped). A key to developing the method is regarding the hidden and dropped transition data as labeled and unlabeled multi-step transition data, where the labels represent the number of steps required for each transition. This allows us to describe the generative process of multi-step transition data, and thus develop a new probabilistic model. We confirm the effectiveness of the proposal by experiments on synthetic and real data.

In this paper, we propose a method that can handle censored data, data collected under the condition that the exact value is recorded only when the value is within a certain range, abbreviated information is recorded otherwise. It is known that existing methods that use mixture models with censored data suffer from (i) the existence of local optimum solutions and (ii) the need to compute the statistics of truncated distributions for parameter estimation. Our proposal, exemplar based censored mixture model (EBCM), overcomes these two difficulties at once by adopting the exemplar based model approach. The effectiveness of EBCM is confirmed by experiments on synthetic and real world dat sets.

In this paper, we propose a probabilistic model for analyzing the generalized interval valued matrix, a matrix that has scalar valued elements and bounded/unbounded interval valued elements. We derive a majorization minimization algorithm for parameter estimation and prove that the objective function is monotonically decreasing by the parameter update. An experiment shows that the proposed model well handles interval- valued elements and offers improved performance.

In this paper, we tackle the problem of extracting latent structure and patterns from multiple datasets that consist of users' rating scores and activity logs (click, view, visit, ...) in order to understand the typical users' behavior. Our proposed method is based on non-negative matrix factorization, and factorizes multiple matrices simultaneously while adopting Euclidean distance and generalized KL divergence for the rating matrix and the activity matrix, respectively. We derive an optimization algorithm that offers a theoretical guarantee that it can find a locally optimal solution. Our experiments show that the proposed method outperformed existing methods when measured by mean squared error, which implies that it can extract latent structure and patterns more precisely. We also confirm that the segmentation result by the proposal helps to analyze users' behavior.

In this paper, we tackle with the problem of analyzing datasets with different resolution such as a pair of user's individual data and user group's data, for example "userA visited shopA 5 times" and "users whose attributes are men purchased itemA 80 times in total". In order to establish a basic approach to this problem, we focus on the simplified scenario and propose a new probabilistic model called probabilistic non-negative inconsistent-resolution matrices factorization (pNimf). pNimf is rigorously derived from the data generative process using latent high-resolution data which underlie low-resolution data. We conduct experiments on real purchase log data and confirm that the proposed model provides superior performance, and that the performance improves as the number of low-resolution data increases. These results imply that our way of modeling using latent high-resolution data can become the basic approach to the problem of analyzing dataset with different resolution.

Recently, analysis of massive purchase histories is thought to be effective for marketing science. For successful sales strategies and advertisements, understanding of consumer's attitude (motivation toward purchasing) is important. However, marketers could obtain merely static attitude for the limited consumers using subjective questionnaires. We focus on purchase rhythm (regularity) as one approach to automatically extracting dynamic attitude transition. We propose "regular behavior measure" which quantifies how well each purchase corresponds to the user's own daily habits. It considers non-periodic habits as well as periodic habits using multiple probabilistic models, we consider purchase probability and the number of purchased dates simultaneously. We also define "influence weight" to quantify whether each item is purchased on a fixed day of the week (DoW) or year-month (YM) using Normalized Mutual Information. We analyze actual purchase data to show the effectiveness of our visualization that considers consumer's regularity transitions.