Deep sum-product networks learn faster than shallow models.
problem The speed of parameter optimization in sum-product networks.
method Theoretical analysis and empirical experiments on overparameterized sum-product networks.
result Gradient-based optimization in deep sum-product networks is equivalent to gradient ascent with adaptive and time-varying learning rates and additional momentum terms.
Production networks amplify economic growth through technology diffusion.
problem Understanding how technology improvements propagate through production networks.
method Analyzing a production network model to study the effects of technological improvements.
result Longer production chains lead to faster price reduction and GDP growth.
Complex products trade through fewer countries, making them more fragile.
problem Fragility in the global economy due to centralized trade networks for complex products.
method Used network science and product complexity theory indicators to analyze trade networks.
result Products with higher complexity trade through fewer countries, making them more fragile.
With globalization, countries are more connected than before by trading flows, which currently amount to at least 36 trillion dollars. Interestingly, approximately 30-60 percent of global exports consist of intermediate products. Therefore, the trade flow network of a particular product with high added values can be re…
Few attempts have been proposed in order to describe the statistical features and historical evolution of the export bipartite matrix countries/products. An important standpoint is the introduction of a products network, namely a hierarchical forest of products that models the formation and the evolution of commodities…
Improved sales forecasting for new products using transfer learning.
problem Insufficient training data for new products leads to inaccurate sales forecasts.
method Network-based Transfer Learning approach for deep neural networks.
result Deep neural networks' prediction accuracy for food sales forecasting can be effectively increased.
Study analyzes Colombian firms' export capabilities over 5 years.
problem Understanding specialization in Colombian firms' export products.
method Bipartite network analysis, modularity maximization, Louvain algorithm.
result Firms specialize in exporting specific product categories, forming clusters.
Paper shows neural network derivatives have outer product structure.
problem Understanding and utilizing higher-order information in neural networks.
method Analyzes feedforward, recurrent, and convolutional neural networks to identify derivative structure.
result Derivatives of feedforward and recurrent networks have an outer product structure, while convolutional networks do not.
New method combines FMEA and Bayesian Network for root cause analysis in lithium-ion battery production.
problem Complex cause-effect relationships in lithium-ion battery production.
method Combining FMEA with Bayesian Network to detect and resolve inconsistencies.
result Holistic method builds large-scale cross-process Bayesian Failure Network for root cause analysis.
A neural network method estimates entropy production from system trajectories.
problem Estimating entropy production from system trajectories without detailed dynamics.
method Developed a neural estimator (NEEP) for entropy production (EP).
result NEEP rigorously proves to provide stochastic EP by optimizing an objective function.
Dynamic pricing learns demand model from sparse product networks.
problem Minimizing revenue loss in a large network of products with unknown demand parameters.
method Combines optimism-in-the-face-of-uncertainty and PAC-Bayesian approaches.
result Achieves asymptotically optimal performance in terms of network size and time horizon.
ABIPNN improves neural network performance by processing vectors in each neuron.
problem Traditional neural networks fail to model associations among adjacent scalars.
method ABIPNN uses arbitrary bilinear products to process vector-valued neurons.
result ABIPNN outperforms conventional neural networks in multispectral image denoising and singing voice separation.
Efficiently compress SPNs using tensor networks.
problem Efficiently compressing Sum-Product Networks (SPNs).
method Mapping SPNs onto tensor networks and employing novel optimization techniques.
result Remarkable parameter compression with negligible loss in accuracy.
We introduce an algorithm able to reconstruct the relevant network structure on which the time evolution of country-product bipartite networks takes place. The significant links are obtained by selecting the largest values of the projected matrix. We first perform a number of tests of this filtering procedure on synthe…
NeuroPMD estimates densities on complex product manifolds.
problem Density estimation on high-dimensional product manifolds.
method Neural network directly parameterizes density, trained with manifold differential operators.
result NeuroPMD outperforms traditional methods in density estimation.
Sum-Product-Quotient Networks boost generative model power by incorporating conditional distributions.
problem Limited expressivity of Sum-Product Networks (SPNs).
method Integrates conditional distributions using quotient nodes and provides tractability conditions.
result Proves SPQNs can compute some distributions more efficiently than SPNs, reducing size requirements.
Modeling how network connectivity affects economic collapse and robustness.
problem Impact of network topology on systemic risk and collapse of complex economic systems.
method Proposed a model to study the effects of network structure on economic systems by varying connectivity.
result Emergent systemic risks arise with increased interconnections, leading to phase transitions and tipping points.
SPTN uses invertible transformations to improve sum-product networks.
problem Improving inference efficiency and tractability in sum-product networks.
method Integrates invertible transformations into sum-product networks (SPNs).
result SPTNs with Gaussian leaves and affine transformations are as tractable as SPNs.
A new neural network model ONCF improves collaborative filtering by explicitly modeling embedding correlations.
problem Improving collaborative filtering for better recommendation quality.
method ONCF uses an outer product to model embedding correlations and a convolutional layer to learn high-order correlations.
result ONCF outperforms existing models in implicit feedback data experiments.
Quantum spin networks' tails are q-series linked to colored Jones polynomials.
problem Understanding the q-series of quantum spin networks. method Analyzing the q-series of quantum spin networks and their relation to the colored Jones polynomial. result Quantum spin networks' tails are related to the tail of the colored Jones polynomial.
We propose a simple dynamical model of the formation of production networks among monopolistically competitive firms. The model subsumes the standard general equilibrium approach à la Arrow-Debreu but displays a wide set of potential dynamic behaviors. It robustly reproduces key stylized facts of firms' demographics. O…
Traditionally, multi-layer neural networks use dot product between the output vector of previous layer and the incoming weight vector as the input to activation function. The result of dot product is unbounded, thus increases the risk of large variance. Large variance of neuron makes the model sensitive to the change o…
Although standard economics textbooks are seldom interested in production networks, modern economies are more and more based upon suppliers/customers interactions. One can consider entire sectors of the economy as generalised supply chains. We will take this view in the present paper and study under which conditions lo…
Paper proposes PNN and PIN models for better user response prediction.
problem User response prediction in multi-field categorical data with sparse representations.
method Kernel product for field-aware feature interactions, Product-based Neural Network (PNN) for DNN-based models, and Product-network In Network (PIN) for generalization.
result PNN and PIN models consistently outperform 8 baselines on AUC and log loss.
The easy access to large data sets has allowed for leveraging methodology in network physics and complexity science to disentangle patterns and processes directly from the data, leading to key insights in the behavior of systems. Here we use to country specific food production data to study binary and weighted topologi…
Proposes MRNet-Product2Vec for product embeddings in e-commerce.
problem Creating dense, low-dimensional product embeddings for diverse tasks.
method Discriminative Multi-task Bidirectional Recurrent Neural Network (RNN) with Bidirectional RNN input and fifteen product labels output.
result Product embeddings perform almost as well as TF-IDF but with less dimensionality.
ProductNet curates high-quality product datasets for better product understanding.
problem Lack of high-quality product datasets for product representation learning.
method Curated high-quality product datasets with a multi-modal deep neural network and active learning.
result Master model yields high categorization accuracy (94.7% top-1 accuracy for 1240 classes).
Improved collaborative filtering with neural network models of reviews.
problem Improve collaborative filtering performance using side information from reviews.
method Introduced two neural network models (product-of-experts and recurrent neural network) to incorporate reviews into collaborative filtering.
result The product-of-experts model achieved state-of-the-art performance, outperforming LDA-based approach.
Proposes a network framework for forecasting futures with different expirations.
problem Forecasting E-mini S\&P 500 and CBOE Volatility Index futures with different expirations.
method A novel data-driven network framework using GCN-LSTM, visualizing correlation structures, and enhancing LSTM's predictive power.
result Enhanced predictive power of future forecasts through a multi-channel Graph Convolutional Network.
Researchers develop a new method to learn from incomplete data.
problem Learning from incomplete data with imprecise probabilities.
method Credal sum-product networks (SPNs) for robust probabilistic representations.
result Imprecise SPNs can capture robustness to missing data.
Novel analysis of neural networks using geometric algebra and convex optimization.
problem Understanding the inner workings of deep neural networks.
method Geometric (Clifford) algebra and convex optimization.
result Optimal weights are given by the wedge product of training samples.
Bayesian networks improve product risk assessment by handling uncertainty and causality.
problem Limited handling of uncertainty and inability to incorporate causal explanations in existing methods.
method Bayesian Networks (BNs) for improved systematic product risk assessment.
result BN approach provides more powerful and flexible risk assessments.
Predicting the future evolution of complex systems is one of the main challenges in complexity science. Based on a current snapshot of a network, link prediction algorithms aim to predict its future evolution. We apply here link prediction algorithms to data on the international trade between countries. This data can b…
WIPS optimizes inner product weights to approximate various similarities.
problem Learning high-quality node representations and accurate similarities.
method Weighted inner product similarity (WIPS) with adjustable weights.
result WIPS can approximate arbitrary general similarities including positive definite and indefinite kernels.
New method uses outer product manifolds to simplify neural networks.
problem Exponential inefficiency of hierarchical neural networks.
method Reparametrization invariant Riemannian metrics and tangent subspace computation.
result Significant improvement in network performance after early training.
Local logarithmic export distributions show non-zero skewness that changes with exporter and destination characteristics.
problem Identifying the skewness in local logarithmic export distributions and its relationship with exporter and destination characteristics.
method Analyzing directed links weighted by the logarithm of export values, studying the skewness of local exports, and formulating quantitative relations.
result Non-zero skewness in local logarithmic export distributions changes with exporter and destination characteristics.
FGNN generalizes graph neural networks to capture higher-order dependencies.
problem Capturing higher-order dependencies in graph-structured data.
method Introducing a factor graph neural network (FGNN) that can represent Max-Product Belief Propagation.
result FGNN effectively represents Max-Product Belief Propagation and performs well on both synthetic and real datasets.
Sum-product networks have recently emerged as an attractive representation due to their dual view as a special type of deep neural network with clear semantics and a special type of probabilistic graphical model for which inference is always tractable. Those properties follow from some conditions (i.e., completeness an…
Tricks improve retail product image classification accuracy.
problem Retail Product Image Classification
method Various tricks including a new LCA layer, Instagram-pretrained Convnet, and Maximum Entropy loss.
result Increased accuracy of fine-tuned convnets by a large margin.
Study characterizes community structure in Japanese production network.
problem Characterize community structure in a large-scale production network.
method Directed network analysis of one million Japanese firms.
result Large fraction of firms have local interactions, and community strengths are heterogeneous.
The paper develops AI methods to predict weather-sensitive product sales.
problem Lack of accurate demand prediction for weather-sensitive products.
method Artificial Neural Networks applied to historical Walmart sales data.
result Improved demand prediction for weather-sensitive products.
Proposes RBGP framework for efficient block sparse neural networks.
problem Efficiently exploit structured sparsity patterns for sparse neural networks on GPU.
method Uses Ramanujan Bipartite Graph Product to generate structured multi-level block sparse neural networks.
result Achieves 5-9x and 2-5x runtime gains over unstructured and block sparsity patterns respectively, while maintaining accuracy.
Sum-Product Networks (SPN) have recently emerged as a new class of tractable probabilistic graphical models. Unlike Bayesian networks and Markov networks where inference may be exponential in the size of the network, inference in SPNs is in time linear in the size of the network. Since SPNs represent distributions over…
Deep neural network predicts product returns before purchase.
problem High costs of handling returned fashion products.
method Bayesian Personalized Ranking (BPR) embeddings and skip-gram model for user and product features.
result Reduced overall returns through real-time return probability prediction.
Model assesses how supply chain disruptions affect financial stability.
problem Systemic risk in production networks and its financial implications.
method Data-driven econo-financial stress-testing framework combining supply chain and interbank networks.
result Increase of up to 28% in financial systemic risk due to production network contagion.
Much of the analysis of economic growth has focused on the study of aggregate output. Here, we deviate from this tradition and look instead at the structure of output embodied in the network connecting countries to the products that they export.We characterize this network using four structural features: the negative r…
Graphical notation simplifies tensor operations and decompositions.
problem Complex tensor operations are difficult to understand and represent.
method Introduces graphical notation to represent tensor operations.
result Simplified representation of tensor operations and decompositions.
Combines gating and tensor products for RNNs to improve performance.
problem Improving RNNs' ability to capture long-term dependencies.
method Proposes a novel RNN architecture combining gating mechanism and tensor products.
result Significant performance improvement on word-level and character-level language modeling tasks.