David I. Inouye

Assistant Professor

School of Electrical and Computer Engineering (ECE)

Biography

I am an assistant professor in the School of Electrical and Computer Engineering at Purdue University. Previously, I was a postdoc at Carnegie Mellon University working with Prof. Pradeep Ravikumar. I completed my Computer Science PhD at The University of Texas at Austin in 2017 advised by Prof. Inderjit Dhillon and Prof. Pradeep Ravikumar. I was awarded the NSF Graduate Research Fellowship (NSF GRFP). My research interests are in machine learning focused on the fundamentals of generative probabilistic models and explainable AI.

Interests

Deep generative models
Invertible deep learning
Explainable and interpretable AI
Distribution shift detection

Education

PostDoc in Machine Learning, 2019

Carnegie Mellon University
PhD in Computer Science, 2017

The University of Texas at Austin
MS in Computer Science, 2015

The University of Texas at Austin
BS in Electrical Engineering, 2012

Georgia Institute of Technology
BA in Natural Sciences, 2011

Covenant College

Selected Publications

(New!) Feature Shift Detection: Localizing Which Features Have Shifted via Conditional Distribution Tests

While previous distribution shift detection approaches can identify if a shift has occurred, these approaches cannot localize which …

Sean M. Kulinski, Saurabh Bagchi, David I. Inouye. Neural Information Processing Systems (NeurIPS) (accepted), 2020.

PDF

(New!) WeakFlow: Iterative Invertible Distribution Transformations via Weak Destructive Flow

David I. Inouye, Pradeep Ravikumar. ICML Workshop on Invertible Neural Networks, Normalizing Flows, and Explicit Likelihood Models (INNF+), 2020.

Automated Dependence Plots

In practical applications of machine learning, it is necessary to look beyond standard metrics such as test accuracy in order to …

David I. Inouye, Liu Leqi, Joon Sik Kim, Bryon Aragam, Pradeep Ravikumar. Uncertainty in Artificial Intelligence (UAI), 2020.

PDF Code

On the (In)fidelity and Sensitivity of Explanations

We consider objective evaluation measures of explanations of complex black-box machine learning models. We propose simple robust …

Chih-Kuan Yeh, Cheng-Yu Hsieh, Arun S. Suggala, David I. Inouye, Pradeep Ravikumar. Neural Information Processing Systems (NeurIPS), 2019.

PDF Code

Deep Point Process Destructors

David I. Inouye. NeurIPS Workshop on Learning with Temporal Point Processes, 2019.

Deep Density Destructors

We propose a unified framework for deep density models by formally defining density destructors. A density destructor is an invertible …

David I. Inouye, Pradeep Ravikumar. International Conference on Machine Learning (ICML), 2018.

PDF Code

Hyperparameter Selection under Localized Label Noise via Corrupt Validation

Existing research on label noise often focuses on simple uniform or classconditional noise. However, in many real-world settings, label …

David I. Inouye, Pradeep Ravikumar, Pradipto Das, Ankur Datta. Learning with Limited Labeled Data (NeurIPS Workshop), 2017.

PDF Code

A Review of Multivariate Distributions for Count Data Derived from the Poisson Distribution

The Poisson distribution has been widely studied and used for modeling univariate count-valued data. Multivariate generalizations of …

David I. Inouye, Eunho Yang, Genevera I. Allen, Pradeep Ravikumar. Wiley Interdisciplinary Reviews: Computational Statistics, 2017.

Preprint PDF Code

Square Root Graphical Models: Multivariate Generalizations of Univariate Exponential Families that Permit Positive Dependencies

Square Root Graphical Models (SQR), a novel class of parametric graphical models that provides multivariate generalizations of …

David I. Inouye, Pradeep Ravikumar, Inderjit S. Dhillon. International Conference on Machine Learning (ICML), 2016.

PDF Poster Code

Fixed-Length Poisson MRF: Adding Dependencies to the Multinomial

We propose a novel distribution that generalizes the Multinomial distribution to enable dependencies between dimensions. Our novel …

David I. Inouye, Pradeep Ravikumar, Inderjit S. Dhillon. Neural Information Processing Systems (NeurIPS), 2015.

PDF Poster Code

Capturing Semantically Meaningful Word Dependencies with an Admixture of Poisson MRFs

We develop a fast algorithm for the Admixture of Poisson MRFs (APM) topic model and propose a novel metric to directly evaluate this …

David I. Inouye, Pradeep Ravikumar, Inderjit S. Dhillon. Neural Information Processing Systems (NeurIPS), 2014.

PDF Poster Code

Admixture of Poisson MRFs: A Topic Model with Word Dependencies

This paper introduces a new topic model based on an admixture of Poisson Markov Random Fields (APM), which can model dependencies …

David I. Inouye, Pradeep Ravikumar, Inderjit S. Dhillon. International Conference on Machine Learning (ICML), 2014.

PDF Poster Slides Code

Other Publications

Summarization of Twitter Microblogs

Owing to the sheer volume of text generated by a microblog site like Twitter, it is often difficult to fully understand what is being …

Beaux P. Sharifi, David I. Inouye, Jugal K. Kalita. Computer Journal, 2014.

PDF DOI

Towards Undistorted and Noise-free Speech in an MRI Scanner: Correlation Subtraction Followed by Spectral Noise Gating

Noise cancellation in an MRI environment is difficult due to the high noise levels that are in the spectral range of human speech. This …

Joshua M. Inouye, Silvia S. Blemker, David I. Inouye. The Journal of the Acoustical Society of America, 2014.

PDF DOI

Comparing Twitter Summarization Algorithms for Multiple Post Summaries

Due to the sheer volume of text generated by a micro log site like Twitter, it is often difficult to fully understand what is being …

David I. Inouye, Jugal K. Kalita. IEEE International Conference on Social Computing (SocialCom), 2011.

PDF DOI

Teaching

(New!) ECE 57000: Artificial Intelligence (Fall 2020)

A graduate-level project-based introduction to artificial intelligence (AI) with a primary focus on unsupervised learning. The lecture …

David I. Inouye

Last updated on Sat, Nov 21, 2020

ECE 20875: Python for Data Science (Spring 2020)

This course introduces Python programming to students through data science problems. Students learn Python concepts as well as …

David I. Inouye, Christopher G. Brinton

Last updated on Thu, May 21, 2020

ECE 57000: Artificial Intelligence (Fall 2019)

A graduate-level project-based introduction to artificial intelligence (AI) with a primary focus on unsupervised learning. The lecture …

David I. Inouye

Last updated on Sat, Nov 2, 2019

David I. Inouye

Assistant Professor

School of Electrical and Computer Engineering (ECE)

Purdue University

EE 332

Biography

Interests

Education

Selected Publications

(New!) Feature Shift Detection: Localizing Which Features Have Shifted via Conditional Distribution Tests

(New!) WeakFlow: Iterative Invertible Distribution Transformations via Weak Destructive Flow

Automated Dependence Plots

On the (In)fidelity and Sensitivity of Explanations

Deep Point Process Destructors

Deep Density Destructors

Hyperparameter Selection under Localized Label Noise via Corrupt Validation

A Review of Multivariate Distributions for Count Data Derived from the Poisson Distribution

Square Root Graphical Models: Multivariate Generalizations of Univariate Exponential Families that Permit Positive Dependencies

Fixed-Length Poisson MRF: Adding Dependencies to the Multinomial

Capturing Semantically Meaningful Word Dependencies with an Admixture of Poisson MRFs

Admixture of Poisson MRFs: A Topic Model with Word Dependencies

Other Publications

Summarization of Twitter Microblogs

Towards Undistorted and Noise-free Speech in an MRI Scanner: Correlation Subtraction Followed by Spectral Noise Gating

Comparing Twitter Summarization Algorithms for Multiple Post Summaries

Teaching

(New!) ECE 57000: Artificial Intelligence (Fall 2020)

ECE 20875: Python for Data Science (Spring 2020)

ECE 57000: Artificial Intelligence (Fall 2019)

Contact