Blog | Machine Learning | Carnegie Mellon University

Explaining a Black-box Using Deep Variational Information Bottleneck Approach

May 17, 2019

The Rise of Artificial Intelligence Over the past decade, artificial intelligence (AI) has achieved remarkable success in many fields such as healthcare, automotive, and marketing. The capabilities of sophisticated, autonomous decision systems driven by AI keep evolving and moving from…

deep learning machine learning

17 1503

Your 2 is My 1, Your 3 is My 9: Handling Crazy Miscalibrations in Ratings from People

May 4, 2019

Consider the following problem: we are given a set of items, and the goal is to pick the “best” ones from them. This problem appears very often in real life — for example, selecting papers in conference peer review, judging…

machine learning

36 1374

Representer Point Selection for Explaining Deep Neural Networks

April 19, 2019

Why did a Deep Neural Network (DNN) make a certain prediction? Although DNNs have been shown to be extremely accurate predictors in a range of domains, they are still largely black-box functions—even to the experts who train them—due to their…

artificial intelligence deep learning machine learning

39 9504

Building Machine Learning Models via Comparisons

March 29, 2019

Nowadays most machine learning (ML) models predict labels from features. In classification tasks, an ML model predicts a categorical value and in regression tasks, an ML model predicts a real value. These ML models thus require a large amount of…

machine learning

55 11268 3

A Continuous-Time View of Early Stopping for Least Squares (or: How I Learned to Stop Worrying and Love Early Stopping)

March 7, 2019

These days, it seems as though tools from statistics and machine learning are being used almost everywhere. Methods from the literature can now be seen on the critical path in a number of different fields and industries. A consequence of…

machine learning

60 2892

Contextual Parameter Generation for Neural Machine Translation

January 14, 2019

Machine translation is the problem of translating sentences from some source language to a target language. Neural machine translation (NMT), directly models the mapping of a source language to a target language without any need for training or tuning any…

deep learning machine learning machine translation natural language processing

75 3099

Massively Parallel Hyperparameter Optimization

December 12, 2018

Machine learning algorithms typically have configuration parameters, or hyperparameters, that influence their output and ultimately predictive accuracy (Melis et al., 2018). Some common examples of hyperparameters include learning rate, dropout, and activation function for neural networks, maximum tree depth…

automl machine learning

106 10264

Introducing the ML@CMU Blog

December 12, 2018

CMU is a leader in the field of machine learning research, both within the Machine Learning Department (MLD) and across the university in general. Traditional conference and journal publications, along with technical talks, are our primary avenues for disseminating our…

machine learning

103 14553