In the user targeting and expanding of new shows on a video platform, the key point is how their embeddings are generated. It's supposed to be personalized from the perspective of both users and shows. Furthermore, the pursue of both instant (click) and long-time (view time) rewards, and the cold-start problem for new shows bring additional challenges. Such a problem is suitable for processing by heterogeneous graph models, because of the natural graph structure of data. But real-world networks usually have billions of nodes and various types of edges. Few existing methods focus on handling large-scale data and exploiting different types of edges, especially the latter. In this paper, we propose a two-stage audience expansion scheme based on an edge-prompted heterogeneous graph network which can take different double-sided interactions and features into account. In the offline stage, to construct the graph, user IDs and specific side information combinations of the shows are chosen to be the nodes, and click/co-click relations and view time are used to build the edges. Embeddings and clustered user groups are then calculated. When new shows arrive, their embeddings and subsequent matching users can be produced within a consistent space. In the online stage, posterior data including click/view users are employed as seeds to look for similar users. The results on the public datasets and our billion-scale data demonstrate the accuracy and efficiency of our approach.

We present a new data analysis perspective to determine variable importance regardless of the underlying learning task. Traditionally, variable selection is considered an important step in supervised learning for both classification and regression problems. The variable selection also becomes critical when costs associated with the data collection and storage are considerably high for cases like remote sensing. Therefore, we propose a new methodology to select important variables from the data by first creating dependency networks among all variables and then ranking them (i.e. nodes) by graph centrality measures. Selecting Top-$n$ variables according to preferred centrality measure will yield a strong candidate subset of variables for further learning tasks e.g. clustering. We present our tool as a Shiny app which is a user-friendly interface development environment. We also extend the user interface for two well-known unsupervised variable selection methods from literature for comparison reasons.

Efficient representation of quantum many-body states on classical computers is a problem of enormous practical interest. An ideal representation of a quantum state combines a succinct characterization informed by the system's structure and symmetries, along with the ability to predict the physical observables of interest. A number of machine learning approaches have been recently used to construct such classical representations [1-6] which enable predictions of observables [7] and account for physical symmetries [8]. However, the structure of a quantum state gets typically lost unless a specialized ansatz is employed based on prior knowledge of the system [9-12]. Moreover, most such approaches give no information about what states are easier to learn in comparison to others. Here, we propose a new generative energy-based representation of quantum many-body states derived from Gibbs distributions used for modeling the thermal states of classical spin systems. Based on the prior information on a family of quantum states, the energy function can be specified by a small number of parameters using an explicit low-degree polynomial or a generic parametric family such as neural nets, and can naturally include the known symmetries of the system. Our results show that such a representation can be efficiently learned from data using exact algorithms in a form that enables the prediction of expectation values of physical observables. Importantly, the structure of the learned energy function provides a natural explanation for the hardness of learning for a given class of quantum states.

The author's recent research papers, "Cumulative deviation of a subpopulation from the full population" and "A graphical method of cumulative differences between two subpopulations" (both published in volume 8 of Springer's open-access "Journal of Big Data" during 2021), propose graphical methods and summary statistics, without extensively calibrating formal significance tests. The summary metrics and methods can measure the calibration of probabilistic predictions and can assess differences in responses between a subpopulation and the full population while controlling for a covariate or score via conditioning on it. These recently published papers construct significance tests based on the scalar summary statistics, but only sketch how to calibrate the attained significance levels (also known as "P-values") for the tests. The present article reviews and synthesizes work spanning many decades in order to detail how to calibrate the P-values. The present paper presents computationally efficient, easily implemented numerical methods for evaluating properly calibrated P-values, together with rigorous mathematical proofs guaranteeing their accuracy, and illustrates and validates the methods with open-source software and numerical examples.

Water crisis is a crucial concern around the globe. Appropriate and timely maintenance of water pumps in drought-hit countries is vital for communities relying on the well. In this paper, we analyze and apply a sequential attentive deep neural architecture, TabNet, for predicting water pump repair status in Tanzania. The model combines the valuable benefits of tree-based algorithms and neural networks, enabling end-to-end training, model interpretability, sparse feature selection, and efficient learning on tabular data. Finally, we compare the performance of TabNet with popular gradient tree-boosting algorithms like XGBoost, LightGBM,CatBoost, and demonstrate how we can further uplift the performance by choosing focal loss as the objective function while training on imbalanced data.

Protecting data privacy is paramount in the fields such as finance, banking, and healthcare. Federated Learning (FL) has attracted widespread attention due to its decentralized, distributed training and the ability to protect the privacy while obtaining a global shared model. However, FL presents challenges such as communication overhead, and limited resource capability. This motivated us to propose a two-stage federated learning approach toward the objective of privacy protection, which is a first-of-its-kind study as follows: (i) During the first stage, the synthetic dataset is generated by employing two different distributions as noise to the vanilla conditional tabular generative adversarial neural network (CTGAN) resulting in modified CTGAN, and (ii) In the second stage, the Federated Probabilistic Neural Network (FedPNN) is developed and employed for building globally shared classification model. We also employed synthetic dataset metrics to check the quality of the generated synthetic dataset. Further, we proposed a meta-clustering algorithm whereby the cluster centers obtained from the clients are clustered at the server for training the global model. Despite PNN being a one-pass learning classifier, its complexity depends on the training data size. Therefore, we employed a modified evolving clustering method (ECM), another one-pass algorithm to cluster the training data thereby increasing the speed further. Moreover, we conducted sensitivity analysis by varying Dthr, a hyperparameter of ECM at the server and client, one at a time. The effectiveness of our approach is validated on four finance and medical datasets.

Generative Adversarial Networks (GANs) have emerged as a significant player in generative modeling by mapping lower-dimensional random noise to higher-dimensional spaces. These networks have been used to generate high-resolution images and 3D objects. The efficient modeling of 3D objects and human faces is crucial in the development process of 3D graphical environments such as games or simulations. 3D GANs are a new type of generative model used for 3D reconstruction, point cloud reconstruction, and 3D semantic scene completion. The choice of distribution for noise is critical as it represents the latent space. Understanding a GAN's latent space is essential for fine-tuning the generated samples, as demonstrated by the morphing of semantically meaningful parts of images. In this work, we explore the latent space and 3D GANs, examine several GAN variants and training methods to gain insights into improving 3D GAN training, and suggest potential future directions for further research.

Training graph neural networks (GNNs) on large graphs is complex and extremely time consuming. This is attributed to overheads caused by sparse matrix multiplication, which are sidestepped when training multi-layer perceptrons (MLPs) with only node features. MLPs, by ignoring graph context, are simple and faster for graph data, however they usually sacrifice prediction accuracy, limiting their applications for graph data. We observe that for most message passing-based GNNs, we can trivially derive an analog MLP (we call this a PeerMLP) with an equivalent weight space, by setting the trainable parameters with the same shapes, making us curious about \textbf{\emph{how do GNNs using weights from a fully trained PeerMLP perform?}} Surprisingly, we find that GNNs initialized with such weights significantly outperform their PeerMLPs, motivating us to use PeerMLP training as a precursor, initialization step to GNN training. To this end, we propose an embarrassingly simple, yet hugely effective initialization method for GNN training acceleration, called MLPInit. Our extensive experiments on multiple large-scale graph datasets with diverse GNN architectures validate that MLPInit can accelerate the training of GNNs (up to 33X speedup on OGB-Products) and often improve prediction performance (e.g., up to $7.97\%$ improvement for GraphSAGE across $7$ datasets for node classification, and up to $17.81\%$ improvement across $4$ datasets for link prediction on metric Hits@10). The code is available at \href{https://github.com/snap-research/MLPInit-for-GNNs}.

Understanding self-supervised learning is important but challenging. Previous theoretical works study the role of pretraining losses, and view neural networks as general black boxes. However, the recent work of Saunshi et al. argues that the model architecture -- a component largely ignored by previous works -- also has significant influences on the downstream performance of self-supervised learning. In this work, we provide the first theoretical analysis of self-supervised learning that incorporates the effect of inductive biases originating from the model class. In particular, we focus on contrastive learning -- a popular self-supervised learning method that is widely used in the vision domain. We show that when the model has limited capacity, contrastive representations would recover certain special clustering structures that are compatible with the model architecture, but ignore many other clustering structures in the data distribution. As a result, our theory can capture the more realistic setting where contrastive representations have much lower dimensionality than the number of clusters in the data distribution. We instantiate our theory on several synthetic data distributions, and provide empirical evidence to support the theory.

The ability of overparameterized deep networks to interpolate noisy data, while at the same time showing good generalization performance, has been recently characterized in terms of the double descent curve for the test error. Common intuition from polynomial regression suggests that overparameterized networks are able to sharply interpolate noisy data, without considerably deviating from the ground-truth signal, thus preserving generalization ability. At present, a precise characterization of the relationship between interpolation and generalization for deep networks is missing. In this work, we quantify sharpness of fit of the training data interpolated by neural network functions, by studying the loss landscape w.r.t. to the input variable locally to each training point, over volumes around cleanly- and noisily-labelled training samples, as we systematically increase the number of model parameters and training epochs. Our findings show that loss sharpness in the input space follows both model- and epoch-wise double descent, with worse peaks observed around noisy labels. While small interpolating models sharply fit both clean and noisy data, large interpolating models express a smooth loss landscape, where noisy targets are predicted over large volumes around training data points, in contrast to existing intuition.