Interest Level Estimation via Multi-Modal Gaussian Process Latent Variable Factorization

The Role of Providers in Mental Health Services Offered to American-Indian Youths

This article presents a novel data-driven approach to nonlinear system control using a behavioral systems framework. A dynamic latent variable autoencoder (DLVAE) is proposed to project the nonlinear physical variable space onto a linear latent variable space. A data-predictive control approach is developed to control the physical process variables through the latent variables. Based on the behavioral systems theory, the proposed data-driven control framework does not require knowledge of the causality of the latent variables. The stability of the controlled system is ensured by utilizing the concept of trajectory-based dissipativity. The robustness of this control approach is achieved by incorporating the Lipschitz bounds between the latent and physical variables under dissipativity conditions.

In recent years, advancements in artificial intelligence, especially in natural language processing (NLP) models, have progressed rapidly. These models demonstrate remarkable results through training on large datasets and extensive architectures. However, the output process is often a black box, and the decision-making process remains unclear. Our research focuses on the internal representations, specifically the latent variables, generated by NLP models. In earlier work, we explored the latent variables and spaces produced by Sentence-BERT using image generation models. This approach aimed to visualize these spaces by converting discrete textual embeddings into images, introducing continuity and revealing novel relationships. This paper presents the development of an image generation model that uses a common decoder for both GPT-2 and Sentence-BERT, aimed at examining how differences in model architecture affect their latent spaces. We also investigate the impact of training dataset differences by comparing models trained in English and Japanese. Our findings indicate that while the models often generate similar outputs, significant differences emerge in sentences containing multiple elements, attributable to the differing focuses and objectives of the models. Our goal is to understand these latent spaces and contribute to the development of explainable AI.

<p><strong>Deep generative models have greatly advanced the field of artificial intelligence by learning the distribution of unlabelled datasets. In this thesis, we aim to develop a novel variational autoencoder (VAE)-based deep generative model that learns meaningful representations for effective unsupervised clustering in the latent space of high-dimensional datasets.</strong></p><p>Existing VAE-based deep clustering algorithms assume a mixture distribution either for the variational posterior over the latent variables or for the prior, to approximate the distribution of unlabelled datasets. The proposed mixtures variational autoencoders (MVAE) model assumes a mixture distribution for both the prior and variational posterior over the latent variables within the VAE framework.</p><p>We apply the variational inference to construct the evidence lower bound (ELBO) of the marginal log-likelihood function for our proposed MVAE model and then derive the analytical form for the clustering assignment estimate, also called the posterior probability estimate. This estimate is associated with each component of the VAE framework and enables the optimization of two distinct ELBO functions based on soft and hard assignment approaches. The training procedure for MVAE involves the joint optimization of both clustering assignment probabilities and model parameters. We evaluate the clustering performance on the latent embeddings of a dataset using both the posterior probability estimate and the Gaussian mixture model (GMM) methods.</p><p>We proposed two variants of the expectation-maximization (EM) algorithm to optimize the parameters for the MVAE model. The MVAE (EM) algorithm provides two distinct, numerically stable solutions for the posterior probability estimate: MVAE(EM-V1), where the estimate is associated with only the prior distribution, and MVAE(EM-V2), where the estimate incorporates both the variational posterior and the prior distribution. Among the proposed models, MVAE(EM-V2) requires less computation time for training. Additionally, this model achieves superior clustering performance on most datasets compared to the baseline models.</p><p>We modified the MVAE(EM-V2) algorithm by including a coefficient (β) with the regularizer term. The trained regularized mixtures VAE model with a small regularization coefficient (β < 1) achieves good unsupervised clustering performance on the test datasets. We further propose a variant of the regularized mixtures VAE model in which the regularization coefficient follows an annealing schedule from β > 1 to β < 1. The scheduled regularized model demonstrates superior clustering accuracy across most benchmark datasets compared to state-of-the-art deep clustering algorithms.</p><p>In this thesis, we point out that assuming a mixture distribution for both the prior and variational posterior components over the latent variables within the VAE framework enhances unsupervised clustering in the latent space. Our proposed models outperform state-of-the-art deep clustering algorithms, including VADE and k-DVAE, as well as the standard VAE model, in cluster analysis of latent representations across most benchmark datasets. Additionally, the proposed models demonstrate reasonable reconstruction performance and generate realistic examples from the latent space.</p>

Recent deep neural network based methods provide accurate binaural source localization performance. These data-driven models map measured binaural cues directly to source locations hence their performance highly depend on the training data distribution. In this paper, we propose a parametric embedding that maps the binaural cues to a low-dimensional space where localization can be done with a nearest-neighbor regression. We implement the embedding using a neural network, optimized to map points that are close to each other in the latent space (the space of source azimuths or elevations) to nearby points in the embedding space, thus the Euclidean distances between the embeddings reflect their source proximities, and the structure of the embeddings forms a manifold, which provides interpretability to the embeddings. We show that the proposed embedding generalizes well in various acoustic conditions (with reverberation) different from those encountered during training, and provides better performance than unsupervised embeddings previously used for binaural localization. In addition, the proposed method performs better than or equally well as a feed-forward neural network based model that directly estimates the source locations from the binaural cues, and it has better results than the feed-forward model when a small amount of training data is used. Moreover, we also compare the proposed embedding using both supervised and weakly supervised learning, and show that in both conditions, the resulting embeddings perform similarly well, but the weakly supervised embedding allows to estimate source azimuth and elevation simultaneously.

Abnormality detection is a one-class classification (OCC) problem where the methods learn either a generative model of the inlier class (e.g., in the variants of kernel principal component analysis) or a decision boundary to encapsulate the inlier class (e.g., in the one-class variants of the support vector machine). Learning schemes for OCC typically train on data solely from the inlier class, but some recent OCC methods have proposed semi-supervised extensions that also leverage a small amount of training data from outlier classes. Other recent methods extend existing principles to employ deep neural network (DNN) models for learning (for the inlier class) either latent-space distributions or autoencoders, but not both. We propose a semi-supervised variational formulation, leveraging generalized-Gaussian (GG) models leading to data-adaptive, robust, and uncertainty-aware distribution modeling in both latent space and image space. We propose a reparameterization for sampling from the latent-space GG to enable backpropagation-based optimization. Results on many publicly available real-world image sets and a synthetic image set show the benefits of our method over existing methods.

Dimension reduction on neural activity paves a way for unsupervised neural decoding by dissociating the measurement of internal neural state repetition from the measurement of external variable tuning. With assumptions only on the smoothness of latent dynamics and of internal tuning curves, the Poisson Gaussian-process latent variable model (P-GPLVM) (Wu et al., 2017) is a powerful tool to discover the low-dimensional latent structure for high-dimensional spike trains. However, when given novel neural data, the original model lacks a method to infer their latent trajectories in the learned latent space, limiting its ability for estimating the internal state repetition. Here, we extend the P-GPLVM to enable the latent variable inference of new data constrained by previously learned smoothness and mapping information. We also describe a principled approach for the constrained latent variable inference for temporally-compressed patterns of activity, such as those found in population burst events (PBEs) during hippocampal sharp-wave ripples, as well as metrics for assessing whether the inferred new latent variables are congruent with a previously learned manifold in the latent space. Applying these approaches to hippocampal ensemble recordings during active maze exploration, we replicate the result that P-GPLVM learns a latent space encoding the animal’s position. We further demonstrate that this latent space can differentiate one maze context from another. By inferring the latent variables of new neural data during running, certain internal neural states are observed to repeat, which is in accordance with the similarity of experiences encoded by its nearby neural trajectories in the training data manifold. Finally, repetition of internal neural states can be estimated for neural activity during PBEs as well, allowing the identification for replay events of versatile behaviors and more general experiences. Thus, our extension of the P-GPLVM framework for unsupervised analysis of neural activity can be used to answer critical questions related to scientific discovery.

Objective. Partial Least Squares (PLS) regression is a suitable linear decoder model for correlated and high dimensional neural data. This algorithm has been widely used in the application of brain–computer interface (BCI) for the decoding of motor parameters. PLS does not consider nonlinear relations between brain signal features. The nonlinear version of PLS that considers a nonlinear relationship between the latent variables has not been proposed for the decoding of intracranial data. This nonlinear model may cause overfitting in some cases due to a larger number of free parameters. In this paper, we develop a new version of nonlinear PLS, namely nonlinear sparse PLS (NLS PLS) and test it in BCI applications. Approach. In motor related BCI systems, improving the decoding accuracy of both kinetic and kinematic parameters of movement is crucial. To do this, two BCI datasets were chosen to decode the force amplitude and position of hand trajectory using the nonlinear and sparse versions of PLS algorithm. In our new NLS PLS method, we considered a polynomial relationship between the latent variables and used the lasso penalization in the latent space to avoid overfitting and to improve the decoding accuracy. Main results. Some linear and nonlinear based PLS models and our new proposed method, NLS PLS, were applied to the two datasets. According to our results, significant improvement from the NLS PLS method is confirmed over other methods. Our results show that nonlinear PLS outperforms generic PLS in the force decoding but it has lower accuracy in the hand trajectory decoding because of high dimensional feature space. By using lasso penalization, we presented a sparse nonlinear PLS-based model that outperforms generic PLS in both datasets and improves the coefficient of determination, 34% in the force decoding and 10% in the hand trajectory decoding. Significance. We constructed a simple PLS-based model that considers a nonlinear relationship between features and it is also robust to overfitting because of using the lasso penalty in the latent space. This model is suitable for a high dimensional and correlated datasets, like intracranial data and can improve the accuracy of estimation.

A deterioration level estimation method via neural network maximizing category-based ordinally supervised multi-view canonical correlation is presented in this paper. This paper focuses on real world data such as industrial applications and has two contributions. First, a novel neural network handling multi-modal features transforms original features into features effectively representing deterioration levels in transmission towers, which are one of the infrastructures, with consideration of only correlation maximization. It can be realized by setting projection matrices maximizing correlations between multiple features into weights of hidden layers. That is, since the proposed network has only a few hidden layers, it can be trained from a small amount of training data. Second, since there exist diverse characteristics and an ordinal scale in deterioration levels, the proposed method newly derives category-based ordinally supervised multi-view canonical correlation analysis (Co-sMVCCA). Co-sMVCCA enables estimation of effective projection considering both within-class divergence and the ordinal scale between classes. Experimental results showed that the proposed method realizes accurate deterioration level estimation.

Generative models rely on the idea that data can be represented in terms of latent variables which are uncorrelated by definition. Lack of correlation among the latent variable support is important because it suggests that the latent-space manifold is simpler to understand and manipulate than the real-space representation. Many types of generative model are used in deep learning, e.g., variational autoencoders (VAEs) and generative adversarial networks (GANs). Based on the idea that the latent space behaves like a vector space Radford et al. (2015), we ask whether we can expand the latent space representation of our data elements in terms of an orthonormal basis set. Here we propose a method to build a set of linearly independent vectors in the latent space of a trained GAN, which we call quasi-eigenvectors. These quasi-eigenvectors have two key properties: i) They span the latent space, ii) A set of these quasi-eigenvectors map to each of the labeled features one-to-one. We show that in the case of the MNIST image data set, while the number of dimensions in latent space is large by design, 98% of the data in real space map to a sub-domain of latent space of dimensionality equal to the number of labels. We then show how the quasi-eigenvectors can be used for Latent Spectral Decomposition (LSD). We apply LSD to denoise MNIST images. Finally, using the quasi-eigenvectors, we construct rotation matrices in latent space which map to feature transformations in real space. Overall, from quasi-eigenvectors we gain insight regarding the latent space topology.

Many failure root cause analysis (RCA) algorithms for microservices have been proposed with the widespread adoption of microservices systems. Existing algorithms generally focus on RCA with ranking single-level (e.g. metric-level or service-level) root cause candidates (RCCs) with comprehensive monitoring metrics. However, many heterogeneous RCCs exist with limited observability in real-world microservices systems. Further, we find that the limited observability may result in inaccurate RCA through real-world failures in eBay. In this paper, for the first time, we propose to "model RCCs as latent variables". The core idea is to infer the status of RCCs as latent variables with related monitoring metrics instead of directly extracting features from only the observable metrics. Based on this, we propose LatentScope, an unsupervised RCA framework with heterogeneous RCCs under limited observability. A dual-space graph is proposed to model both observable and unobservable variables, with many-to-many relationships between spaces. To achieve fast inference of latent variables and RCA, we propose the LatentRegressor algorithm, which includes Regression-based Latent-space Intervention Recognition (RLIR) to achieve intervention recognition-based RCA in latent space. LatentScope has been deployed in eBay's production environment and evaluated on both eBay's real-world failures and a testbed dataset. The evaluation results show that, compared with baseline algorithms, our model significantly improves the Top-1 recall by 9.7%-57.9%. The source code of LatentScope and the dataset are available at https://github.com/NetManAIOps/LatentScope.

Retrosynthesis aims to identify sets of reactants capable of synthesizing a target molecule and has recently benefited from advancements in template-free sequence-translation models, which offer both efficiency and high predictive accuracy. A challenge in this domain is effectively capturing the intrinsic one-to-many relationship characteristic of chemical reactions. To address this, we propose a Hierarchical Conditional Variational Auto-Encoder (HCVAE) module that can be seamlessly integrated into existing template-free retrosynthesis frameworks. Our method establishes a hierarchical latent space that transitions from continuous to discrete representations: a continuous latent variable explores diverse chemical transformation proposals, while a discrete latent variable groups them into high-level reaction classes. This design links one product to multiple possible reactants, thereby enhancing coverage of multicandidate synthesis schemes. Extensive evaluations conducted on three publicly available benchmarks, encompassing both single-step prediction and multistep planning tasks, demonstrate that the HCVAE consistently improves performance across various backbone architectures. For instance, the single-step RootAligned model exhibits an increase in top-10 exact match accuracy on the USPTO-50k data set from 90.5% to 91.6%, meanwhile the DirectMultistep model shows improvements from 49.3% to 53.1% and from 43.0% to 46.7% on the n1 and n5 sets of the PaRoutes data set, respectively. Further analyses indicate that the learned latent space organization provides a structured mechanism for navigating alternative reaction proposals and facilitates practical multistep synthesis of drug-like molecules.

Mining high dimensional data-sets extracted from real world problems is a challenging task due to the large features' space. The latent variables are used to reduce the dimensions of this space by representing highly dependent features. They simplify the creation of probabilistic models and they clarify the semantic of the inferred knowledge. Learning these variables for Bayesian network, as the most generic probabilistic model, is problematic. Actually, there is not a direct way that leads to finding their cardinalities. The precision of the inferred model is highly dependent on the accuracy of the latent variable's cardinality. Therefore, choosing a small value leads to a generalized model having a high rate of information loss. Moreover, a high cardinality tend to over-fit the data, to generate complex latent variables and to burden the parameter learning of the probabilistic model. In this paper, we propose a new criterion based on the mutual information and the log likelihood, called the equilibrium criterion. We mathematically and experimentally validate its efficiency for estimating the cardinality of the latent variable. We also demonstrate its performance in finding the hidden cause of a set of observed variables. The experimental analysis shows that our method succeeded in restoring the original cardinality of intentionally deleted variables in known networks.

Deep directed generative models have attracted much attention recently due to their generative modeling nature and powerful data representation ability. In this article, we review different structures of deep directed generative models and the learning and inference algorithms associated with the structures. We focus on a specific structure that consists of layers of Bayesian networks (BNs) due to the property of capturing inherent and rich dependencies among latent variables. The major difficulty of learning and inference with deep directed models with many latent variables is the intractable inference due to the dependencies among the latent variables and the exponential number of latent variable configurations. Current solutions use variational methods, often through an auxiliary network, to approximate the posterior probability inference. In contrast, inference can also be performed directly without using any auxiliary network to maximally preserve the dependencies among the latent variables. Specifically, by exploiting the sparse representation with the latent space, max-max instead of maxsum operation can be used to overcome the exponential number of latent configurations. Furthermore, the max-max operation and augmented coordinate ascent (AugCA) are applied to both supervised and unsupervised learning as well as to various inference. Quantitative evaluations on benchmark data sets of different models are given for both data representation and feature-learning tasks.

In this paper, a new method for fuzzy clustering is proposed that combines generative topographic mapping (GTM) and Fuzzy c‐means (FCM) clustering. GTM is used to generate latent variables and their posterior probabilities. These two provide the distribution of the input data in the latent space. FCM determines the seeds of clusters, as well as the resultant clusters and the corresponding membership functions of the input data, based on the latent variables obtained from GTM. Experiments are conducted to compare the results obtained using FCM and the Gustafson‐Kessel (GK) algorithm with the proposed method in terms of four cluster‐validity indexes. Using simulated and benchmark data sets, it is observed that the hybrid method (GTMFCM) performs better than FCM and GK algorithms in terms of these indexes. It is also found that the superiority of GTMFCM over FCM and GK algorithms becomes more pronounced with the increase in the dimensionality of the input data set.