Deep Learning on Graphs

An interesting research problem to find a representation of the graph that can be utilized by a machine learning model. Instead of the global summary attributes of the graph, finding a dense representation of the graph that incorporates the local structural information. The idea is to learn a mapping from the graph to a low- dimensional real vector space such that the structural attributes of the graph translate to the geometrical properties in the embedding space. Neural network models have been very succesful at this task. Graph representation learning on heterogeneous networks offers many difficulties, and is not obvious how one can build en embedding that uses the heterogeniety. We propose a method to mine net- work representation that effectively exploits edge heterogeneity in multi-graphs. As an application, we studied the friend recommendation in social networks which is modeled on link prediction in the network.

Visual analysis of clustering

Clustering is a common type of unsupervised machine learning that can be useful for summarizing and aggregating complex multi-dimensional data. However, data can be clustered in many ways, and there exist a large body of algorithms designed to reveal different patterns. In practice, it is quite difficult for data scientists to choose and parameterize algorithms to get the clustering results relevant for their dataset and analytical tasks. To alleviate this problem, we built a clustering analysis system that clusters data using a variety of clustering techniques and parameters and then recommends good clustering results utilizing a variety of quality scoring metrics. Our visual user interface allows users to find high quality clustering results, explore the clusters using a variety of coordinated visualization techniques, and select the cluster result that best suits their task.

Event Sequence Mining and Visualization

Exploring event sequences in big data is challenging. Though many mining algorithms have been developed to derive the most frequently occurring and the most meaningful sequential patterns, it is yet difficult to make sense of the results. To tackle the problem, we introduce a visual analytics approach - Peekquence, which aims to increase the interpretability of machine learning-based sequence mining algorithms.

Biological Pathway Analysis

Biological interpretation and understanding of machine learning based predictive models are highly desirable in healthcare analytics. Predicting Adverse Drug Reactions (ADRs) is extremely important for safe and precision medicine. There are various machine learning based approaches to predict adverse reactions for drugs. These models, though effective, lack biological interpretation and are treated as black-boxes. We propose DrugPathSeeker, a novel in- teractive user interface that integrates the machine learning model, database query API, statistical analysis, and visualization for ex- ploring and understanding of the association between drugs and ADRs.

Weighted Mixed Linear Model for GWAS

Linear mixed models are widely used in humans, animals, and plants to conduct genome-wide association studies (GWAS). Two-stage methods have been proposed to reduce the complexity and increase the computational speed of whole-genome scans. The first stage of the analysis fits raw data to a model including environment and line effects, but no individual marker effects. The second stage involves the whole genome scan of marker tests using summary values for each line as the dependent variable. Missing data and unbalanced experimental designs can result in biased estimates of marker association effects from two-stage analyses. In this study, we developed a weighted two-stage analysis to reduce bias and improve power of GWAS while maintaining the computational efficiency of two-stage analyses.