Timezone: »

A deep learning approach to recover conditional independence graphs
Harsh Shrivastava · Urszula Chajewska · Robin Abraham · Xinshi Chen

Probabilistic Graphical Models are generative models of complex systems. They rely on conditional independence assumptions between variables to learn sparse representations which can be visualized in a form of a graph. Such models are used for domain exploration and structure discovery in poorly understood domains. This work introduces a novel technique to perform sparse graph recovery by optimizing deep unrolled networks. Assuming that the input data $X\in\mathbb{R}^{M\times D}$ comes from an underlying multivariate Gaussian distribution, we apply a deep model on $X$ that outputs the precision matrix $\Theta$. Then, the partial correlation matrix $\rho$' is calculated which can also be interpreted as the conditional independence graph. Our model, \texttt{uGLAD}, builds upon and extends the state-of-the-art model \texttt{GLAD} to the unsupervised setting. The key benefits of our model are (1) \texttt{uGLAD} automatically optimizes sparsity-related regularization parameters leading to better performance than existing algorithms. (2) We introduce multi-task learning based consensus' strategy for robust handling of missing data in an unsupervised setting. We evaluate performance on synthetic Gaussian, non-Gaussian data generated from Gene Regulatory Networks, and present a case study in anaerobic digestion.

#### Author Information

##### Harsh Shrivastava (Microsoft Research, Redmond)

Harsh Shrivastava is a Senior Applied Scientist in the AI Architecture and Strategy team at [Microsoft Research](https://www.microsoft.com/en-us/research/people/hshrivastava/). His research interests include working with probabilistic graphical models, using inductive biases to design deep learning architectures & high performance computing. Harsh holds a PhD in Machine Learning (CSE dept.) from Georgia Institute of Technology. He has an undergraduate (dual degree) in Electronics & ECE with Masters in VLSI circuits from Indian Institute of Technology Kharagpur.