Despite many years of research into latent Dirichlet allocation (LDA), applying LDA to collections of non-categorical items is still challenging. Yet many problems with much richer data share a similar structure and could benefit from the vast literature on LDA. We propose logistic LDA, a novel discriminative variant of latent Dirichlet allocation which is easy to apply to arbitrary inputs. In particular, our model can easily be applied to groups of images, arbitrary text embeddings, and integrate well with deep neural networks. Although it is a discriminative model, we show that logistic LDA can learn from unlabeled data in an unsupervised manner by exploiting the group structure present in the data. In contrast to other recent topic models designed to handle arbitrary inputs, our model does not sacrifice the interpretability and principled motivation of LDA.

Machine learning and data mining techniques have been used extensively in order to detect credit card frauds. However, most studies consider credit card transactions as isolated events and not as a sequence of transactions. In this framework, we model a sequence of credit card transactions from three different perspectives, namely (i) The sequence contains or doesn't contain a fraud (ii) The sequence is obtained by fixing the card-holder or the payment terminal (iii) It is a sequence of spent amount or of elapsed time between the current and previous transactions. Combinations of the three binary perspectives give eight sets of sequences from the (training) set of transactions. Each one of these sequences is modelled with a Hidden Markov Model (HMM). Each HMM associates a likelihood to a transaction given its sequence of previous transactions. These likelihoods are used as additional features in a Random Forest classifier for fraud detection. Our multiple perspectives HMM-based approach offers automated feature engineering to model temporal correlations so as to improve the effectiveness of the classification task and allows for an increase in the detection of fraudulent transactions when combined with the state of the art expert based feature engineering strategy for credit card fraud detection. In extension to previous works, we show that this approach goes beyond ecommerce transactions and provides a robust feature engineering over different datasets, hyperparameters and classifiers. Moreover, we compare strategies to deal with structural missing values.

Knowing when a graphical model is perfect to a distribution is essential in order to relate separation in the graph to conditional independence in the distribution, and this is particularly important when performing inference from data. When the model is perfect, there is a one-to-one correspondence between conditional independence statements in the distribution and separation statements in the graph. Previous work has shown that almost all models based on linear directed acyclic graphs as well as Gaussian chain graphs are perfect, the latter of which subsumes Gaussian graphical models (i.e., the undirected Gaussian models) as a special case. However, the complexity of chain graph models leads to a proof of this result which is indirect and mired by the complications of parameterizing this general class. In this paper, we directly approach the problem of perfectness for the Gaussian graphical models, and provide a new proof, via a more transparent parametrization, that almost all such models are perfect. Our approach is based on, and substantially extends, a construction of Ln\v{e}ni\v{c}ka and Mat\'u\v{s} showing the existence of a perfect Gaussian distribution for any graph.

Opioid overdose rates have reached an epidemic level and state-level policy innovations have followed suit in an effort to prevent overdose deaths. State-level drug law is a set of policies that may reinforce or undermine each other, and analysts have a limited set of tools for handling the policy collinearity using statistical methods. This paper uses a machine learning method called hierarchical clustering to empirically generate "policy bundles" by grouping states with similar sets of policies in force at a given time together for analysis in a 50-state, 10-year interrupted time series regression with drug overdose deaths as the dependent variable. Policy clusters were generated from 138 binomial variables observed by state and year from the Prescription Drug Abuse Policy System. Clustering reduced the policies to a set of 10 bundles. The approach allows for ranking of the relative effect of different bundles and is a tool to recommend those most likely to succeed. This study shows that a set of policies balancing Medication Assisted Treatment, Naloxone Access, Good Samaritan Laws, Medication Assisted Treatment, Prescription Drug Monitoring Programs and legalization of medical marijuana leads to a reduced number of overdose deaths, but not until its second year in force.

One of the most effective algorithms for differentially private learning and optimization is objective perturbation. This technique augments a given optimization problem (e.g. deriving from an ERM problem) with a random linear term, and then exactly solves it. However, to date, analyses of this approach crucially rely on the convexity and smoothness of the objective function. We give two algorithms that extend this approach substantially. The first algorithm requires nothing except boundedness of the loss function, and operates over a discrete domain. Its privacy and accuracy guarantees hold even without assuming convexity. The second algorithm operates over a continuous domain and requires only that the loss function be bounded and Lipschitz in its continuous parameter. Its privacy analysis does not even require convexity. Its accuracy analysis does require convexity, but does not require second order conditions like smoothness. We complement our theoretical results with an empirical evaluation of the non-convex case, in which we use an integer program solver as our optimization oracle. We find that for the problem of learning linear classifiers, directly optimizing for 0/1 loss using our approach can out-perform the more standard approach of privately optimizing a convex-surrogate loss function on the Adult dataset.

Abstract--This paper studies the problem of cross-network node classification to overcome the insufficiency of labeled data in a single network. It aims to leverage the label information in a partially labeled source network to assist node classification in a completely unlabeled or partially labeled target network. Existing methods for single network learning cannot solve this problem due to the domain shift across networks. Some multi-network learning methods heavily rely on the existence of cross-network connections, thus are inapplicable for this problem. T o tackle this problem, we propose a novel network transfer learning framework AdaGCN by leveraging the techniques of adversarial domain adaptation and graph convolution. It consists of two components: a semi-supervised learning component and an adversarial domain adaptation component. The former aims to learn class discriminative node representations with given label information of the source and target networks, while the latter contributes to mitigating the distribution divergence between the source and target domains to facilitate knowledge transfer. Extensive empirical evaluations on real-world datasets show that AdaGCN can successfully transfer class information with a low label rate on the source network and a substantial divergence between the source and target domains. Codes will be released upon acceptance. It is an important building block of numerous real-world applications, such as product recommendation in e-commerce websites, advertisement distribution in social networks, and protein function identification for disease diagnosis. Many research efforts have been made to develop reliable and efficient methods for node classification in networked data. In the era of big data, massive amount of raw data in information networks is produced everyday . However, labeled data is significantly expensive and slow to acquire due to the high cost and long time of human annotations, making it difficult to train a well-generalized classifier [2]. Moreover, in some newly-formed networks such as a protein-protein interaction network constructed by some researchers, there may be no labels at all. Hence, it would be impossible to classify the nodes with only the information of this network. T o tackle these issues, a promising approach is to utilize class information from other similar or related networks to assist in classification, i.e., transfer learning on networked data [3], [4].

Regularization is a popular technique in machine learning for model estimation and avoiding overfitting. Prior studies have found that modern ordered regularization can be more effective in handling highly correlated, high-dimensional data than traditional regularization. The reason stems from the fact that the ordered regularization can reject irrelevant variables and yield an accurate estimation of the parameters. How to scale up the ordered regularization problems when facing the large-scale training data remains an unanswered question. This paper explores the problem of parameter estimation with the ordered $\ell_{2}$-regularization via Alternating Direction Method of Multipliers (ADMM), called ADMM-O$\ell_{2}$. The advantages of ADMM-O$\ell_{2}$ include (i) scaling up the ordered $\ell_{2}$ to a large-scale dataset, (ii) predicting parameters correctly by excluding irrelevant variables automatically, and (iii) having a fast convergence rate. Experiment results on both synthetic data and real data indicate that ADMM-O$\ell_{2}$ can perform better than or comparable to several state-of-the-art baselines.

Link prediction is an important way to complete knowledge graphs (KGs), while embedding-based methods, effective for link prediction in KGs, perform poorly on relations that only have a few associative triples. In this work, we propose a Meta Relational Learning (MetaR) framework to do the common but challenging few-shot link prediction in KGs, namely predicting new triples about a relation by only observing a few associative triples. W e solve few-shot link prediction by focusing on transferring relation-specific meta information to make model learn the most important knowledge and learn faster, corresponding to relation meta and gradient meta respectively in MetaR. Empirically, our model achieves state-of-the-art results on few-shot link prediction KG benchmarks.

Nearest neighbor is a popular class of classification methods with many desirable properties. For a large data set which cannot be loaded into the memory of a single machine due to computation, communication, privacy, or ownership limitations, we consider the divide and conquer scheme: the entire data set is divided into small subsamples, on which nearest neighbor predictions are made, and then a final decision is reached by aggregating the predictions on subsamples by majority voting. We name this method the big Nearest Neighbor (bigNN) classifier, and provide its rates of convergence under minimal assumptions, in terms of both the excess risk and the classification instability, which are proven to be the same rates as the oracle nearest neighbor classifier and cannot be improved. To significantly reduce the prediction time that is required for achieving the optimal rate, we also consider the pre-training acceleration technique applied to the bigNN method, with proven convergence rate. We find that in the distributed setting, the optimal choice of the neighbor k should scale with both the total sample size and the number of partitions, and there is a theoretical upper limit for the latter. Numerical studies have verified the theoretical findings.

The uncertainties in future Bitcoin price make it difficult to accurately predict the price of Bitcoin. Accurately predicting the price for Bitcoin is therefore important for decision-making process of investors and market players in the cryptocurrency market. Using historical data from 01/01/2012 to 16/08/2019, machine learning techniques (Generalized linear model via penalized maximum likelihood, random forest, support vector regression with linear kernel, and stacking ensemble) were used to forecast the price of Bitcoin. The prediction models employed key and high dimensional technical indicators as the predictors. The performance of these techniques were evaluated using mean absolute percentage error (MAPE), root mean square error (RMSE), mean absolute error (MAE), and coefficient of determination (R-squared). The performance metrics revealed that the stacking ensemble model with two base learner (random forest and generalized linear model via penalized maximum likelihood) and support vector regression with linear kernel as meta-learner was the optimal model for forecasting Bitcoin price. The MAPE, RMSE, MAE, and R-squared values for the stacking ensemble model were 0.0191%, 15.5331 USD, 124.5508 USD, and 0.9967 respectively. These values show a high degree of reliability in predicting the price of Bitcoin using the stacking ensemble model. Accurately predicting the future price of Bitcoin will yield significant returns for investors and market players in the cryptocurrency market.