ICML Best Papers CollectingICML stands for International Conference on Machine Learning. ICML has now grown into an annual top-level international conference on machine learning hosted by the International Society for Machine Learning (IMLS).

The most well known drawback of Gaussian processes regression is the computational cost of the exact calculation of these quantities, which scales as O N 3 in time and O N 2 in memory where N is the number of training examples

We argue that without a careful model of delayed outcomes, we cannot foresee the impact a fairness criterion would have if enforced as a constraint on a classification system

As currently there does not seem to exist a reliable strategy to choose hyperparameters in the unsupervised learning of disentangled representations, we argue that future work should make the role of inductive biases and implicit and explicit supervision more explicit

We have discussed a variety of applications, from creating training-set attacks to debugging models and fixing datasets. Underlying each of these applications is a common tool, the influence function, which is based on a simple idea — we can better understand model behavior by lo...

Arbitrary manipulations of the latent Dirichlet allocation model have to be introduced in order to adapt the topic model to a particular context; Zhao et al; Carman et al )

In this paper we address the following question: “Can we approximately sample from a Bayesian posterior distribution if we are only allowed to touch a small mini-batch of data-items for every sample we generate?”

We discuss how the runtime of SVM optimization should decrease as the size of the training data increases. We present theoretical and empirical results demonstrating how a simple subgradient descent approach indeed displays such behavior, at least for linear kernels.

international conference on machine learning, pp.209-216, (2007)

In this paper, we present an information-theoretic approach to learning a Mahalanobis distance function. We formulate the problem as that of minimizing the differential relative entropy between two multivariate Gaussians under con- straints on the distance function. We express th...

We described two non-convex algorithms using ConCave Convex Procedure that bring marked scalability improvements over the corresponding convex approaches, namely for Support Vector Machines and TSVMs

We present a Support Vector Method that can directly optimize a large class of performance measures like F1-score, Precision/Recall Breakeven Point, Precision at k, and ROCArea

In order to demonstrate the computational complexity of incremental least-squares temporal difference learning algorithm, results were conducted with three different problem sizes: 14, 102 and 402 states