Abstract--This letter proposes a novel communication-efficient and privacy-preserving distributed machine learning framework, coined Mix2FLD. To address uplink-downlink capacity asymmetry, local model outputs are uploaded to a server in the uplink as in federated distillation (FD), whereas global model parameters are downloaded in the downlink as in federated learning (FL). This requires a model output-to-parameter conversion at the server, after collecting additional data samples from devices. Index Terms--Distributed machine learning, on-device learning, federated learning, federated distillation, uplink-downlink asymmetry. Federated learning (FL) is a compelling depicted in Figure 1, Mix2FLD is built upon two key algorithms: solution that collectively trains on-device ML models using federated learning after distillation (FLD) [8] and Mixup their local private data [2], [3].

We study the risk (i.e. generalization error) of Kernel Ridge Regression (KRR) for a kernel $K$ with ridge $\lambda>0$ and i.i.d. observations. For this, we introduce two objects: the Signal Capture Threshold (SCT) and the Kernel Alignment Risk Estimator (KARE). The SCT $\vartheta_{K,\lambda}$ is a function of the data distribution: it can be used to identify the components of the data that the KRR predictor captures, and to approximate the (expected) KRR risk. This then leads to a KRR risk approximation by the KARE $\rho_{K, \lambda}$, an explicit function of the training data, agnostic of the true data distribution. We phrase the regression problem in a functional setting. The key results then follow from a finite-size analysis of the Stieltjes transform of general Wishart random matrices. Under a natural universality assumption (that the KRR moments depend asymptotically on the first two moments of the observations) we capture the mean and variance of the KRR predictor. We numerically investigate our findings on the Higgs and MNIST datasets for various classical kernels: the KARE gives an excellent approximation of the risk, thus supporting our universality assumption. Using the KARE, one can compare choices of Kernels and hyperparameters directly from the training set. The KARE thus provides a promising data-dependent procedure to select Kernels that generalize well.

When deploying a deep neural network on constrained hardware, it is possible to replace the network's standard convolutions with grouped convolutions. This allows for substantial memory savings with minimal loss of accuracy. However, current implementations of grouped convolutions in modern deep learning frameworks are far from performing optimally in terms of speed. In this paper we propose Grouped Spatial Pack Convolutions (GSPC), a new implementation of grouped convolutions that outperforms existing solutions. We implement GSPC in TVM, which provides state-of-the-art performance on edge devices. We analyze a set of networks utilizing different types of grouped convolutions and evaluate their performance in terms of inference time on several edge devices. We observe that our new implementation scales well with the number of groups and provides the best inference times in all settings, improving the existing implementations of grouped convolutions in TVM, PyTorch and TensorFlow Lite by 3.4x, 8x and 4x on average respectively. Code is available at https://github.com/gecLAB/tvm-GSPC/

Despite their popularity, to date, the application of normalizing flows on categorical data stays limited. The current practice of using dequantization to map discrete data to a continuous space is inapplicable as categorical data has no intrinsic order. Instead, categorical data have complex and latent relations that must be inferred, like the synonymy between words. In this paper, we investigate Categorical Normalizing Flows, that is normalizing flows for categorical data. By casting the encoding of categorical data in continuous space as a variational inference problem, we jointly optimize the continuous representation and the model likelihood. To maintain unique decoding, we learn a partitioning of the latent space by factorizing the posterior. Meanwhile, the complex relations between the categorical variables are learned by the ensuing normalizing flow, thus maintaining a close-to exact likelihood estimate and making it possible to scale up to a large number of categories. Based on Categorical Normalizing Flows, we propose GraphCNF a permutation-invariant generative model on graphs, outperforming both one-shot and autoregressive flow-based state-of-the-art on molecule generation.

Longitudinal datasets measured repeatedly over time from individual subjects, arise in many biomedical, psychological, social, and other studies. Such multivariate time-series are often high-dimensional and contain missing values. A common approach to analyse this kind of data is to learn a low-dimensional representation using variational autoencoders (VAEs). However, standard VAEs assume that the learned representations are i.i.d., and fail to capture the correlations between the data samples. We propose a novel deep generative model, Longitudinal VAE (L-VAE), that uses a multi-output additive Gaussian process (GP) prior to extend the VAE's capability to learn structured low-dimensional representations imposed by auxiliary covariate information, and also derive a new divergence upper bound for such GPs. Our approach can simultaneously accommodate both time-varying shared and random effects, produce structured low-dimensional representations, disentangle effects of individual covariates or their interactions, and achieve highly accurate predictive performance. We compare our model against previous methods on synthetic and clinical datasets, and demonstrate the state-of-the-art performance in data imputation, reconstruction, and long-term prediction tasks.

Multi-task learning has gained popularity due to the advantages it provides with respect to resource usage and performance. Nonetheless, the joint optimization of parameters with respect to multiple tasks remains an active research topic. Sub-partitioning the parameters between different tasks has proven to be an efficient way to relax the optimization constraints over the shared weights, may the partitions be disjoint or overlapping. However, one drawback of this approach is that it can weaken the inductive bias generally set up by the joint task optimization. In this work, we present a novel way to partition the parameter space without weakening the inductive bias. Specifically, we propose Maximum Roaming, a method inspired by dropout that randomly varies the parameter partitioning, while forcing them to visit as many tasks as possible at a regulated frequency, so that the network fully adapts to each update. We study the properties of our method through experiments on a variety of visual multi-task data sets. Experimental results suggest that the regularization brought by roaming has more impact on performance than usual partitioning optimization strategies. The overall method is flexible, easily applicable, provides superior regularization and consistently achieves improved performances compared to recent multi-task learning formulations.

Google app market captures the school of thought of users via ratings and text reviews. The critique's viewpoint regarding an app is proportional to their satisfaction level. Consequently, this helps other users to gain insights before downloading or purchasing the apps. The potential information from the reviews can't be extracted manually, due to its exponential growth. Sentiment analysis, by machine learning algorithms employing NLP, is used to explicitly uncover and interpret the emotions. This study aims to perform the sentiment classification of the app reviews and identify the university students' behavior towards the app market. We applied machine learning algorithms using the TF-IDF text representation scheme and the performance was evaluated on the ensemble learning method. Our model was trained on Google reviews and tested on students' reviews. SVM recorded the maximum accuracy(93.37\%), F-score(0.88) on tri-gram + TF-IDF scheme. Bagging enhanced the performance of LR and NB with accuracy of 87.80\% and 85.5\% respectively.

In this paper, we study high-dimensional estimation from truncated samples. We focus on two fundamental and classical problems: (i) inference of sparse Gaussian graphical models and (ii) support recovery of sparse linear models. (i) For Gaussian graphical models, suppose $d$-dimensional samples ${\bf x}$ are generated from a Gaussian $N(\mu,\Sigma)$ and observed only if they belong to a subset $S \subseteq \mathbb{R}^d$. We show that ${\mu}$ and ${\Sigma}$ can be estimated with error $\epsilon$ in the Frobenius norm, using $\tilde{O}\left(\frac{\textrm{nz}({\Sigma}^{-1})}{\epsilon^2}\right)$ samples from a truncated $\mathcal{N}({\mu},{\Sigma})$ and having access to a membership oracle for $S$. The set $S$ is assumed to have non-trivial measure under the unknown distribution but is otherwise arbitrary. (ii) For sparse linear regression, suppose samples $({\bf x},y)$ are generated where $y = {\bf x}^\top{{\Omega}^*} + \mathcal{N}(0,1)$ and $({\bf x}, y)$ is seen only if $y$ belongs to a truncation set $S \subseteq \mathbb{R}$. We consider the case that ${\Omega}^*$ is sparse with a support set of size $k$. Our main result is to establish precise conditions on the problem dimension $d$, the support size $k$, the number of observations $n$, and properties of the samples and the truncation that are sufficient to recover the support of ${\Omega}^*$. Specifically, we show that under some mild assumptions, only $O(k^2 \log d)$ samples are needed to estimate ${\Omega}^*$ in the $\ell_\infty$-norm up to a bounded error. For both problems, our estimator minimizes the sum of the finite population negative log-likelihood function and an $\ell_1$-regularization term.

The literature in this domain has so far divided the study of Ĉ into (i) a first exploration of conditions for its existence and uniqueness as a deterministic solution to (1) (e.g., [4, 8, 12]) and (ii) an independent analysis of its statistical properties when seen as a random object (in the large n regime [1] or in the large n, p regime [2, 14]). In the present article, we claim that the study of the conditions of existence (i) and statistical behavior (ii) of Ĉ can be conveniently carried out jointly. Specifically, by means of a flexible framework based on concentration of measure theory and on a new stable semimetric argument, we simultaneously explore the existence and large dimensional ( n, p large) spectral properties of Ĉ . Our findings may be summarized as the following three main contributions to robust statistics and more generally to large dimensional statistics.

This paper proposes FREEtree, a tree-based method for high dimensional longitudinal data with correlated features. Popular machine learning approaches, like Random Forests, commonly used for variable selection do not perform well when there are correlated features and do not account for data observed over time. FREEtree deals with longitudinal data by using a piecewise random effects model. It also exploits the network structure of the features by first clustering them using weighted correlation network analysis, namely WGCNA. It then conducts a screening step within each cluster of features and a selection step among the surviving features, that provides a relatively unbiased way to select features. By using dominant principle components as regression variables at each leaf and the original features as splitting variables at splitting nodes, FREEtree maintains its interpretability and improves its computational efficiency. The simulation results show that FREEtree outperforms other tree-based methods in terms of prediction accuracy, feature selection accuracy, as well as the ability to recover the underlying structure.