Miguel de Benito Delgado

After several years working as a software developer, Miguel pursued studies in pure mathematics in Madrid and Munich. After finishing his PhD in analysis, and a short research stay in ML, he finally transitioned into the field and ended up working as an applied researcher at the appliedAI Initiative, where he went on to found and head the TransferLab.

Miguel's work

Applications of data valuation in machine learning

At TransferLab we have extensively covered existing and developing methods for Data valuation, the task of attributing value to samples in a …

Data-OOB: Out-of-bag Estimate as a Simple and Efficient Data Value

The out-of-bag (OOB) error estimate is a scalable approach to data valuation. Unlike marginal contribution methods, Data-OOB can leverage …

The debate on the accuracy-interpretability tradeoff

In this talk we delve into the gripping debate on the alleged trade-off between accuracy and interpretability. We’ll discuss the literature …

Uncertainty quantification with conformal prediction

Conformal prediction is a method for uncertainty quantification that is based on the idea of conformal sets (CS). A CS is a set of …

CS-Shapley: Class-wise Shapley Values for Data Valuation in Classification

Using in-class accuracy to up-weight the value of a data point and out-of-class accuracy as a discounting factor, the authors define a new …

Data Banzhaf: A Robust Data Valuation Framework for Machine Learning

A definition of data value with much higher rank stability than Shapley values, proven to be most robust among all semi-values, mitigates …

Beta Shapley: a Unified and Noise-reduced Data Valuation Framework for Machine Learning

In data valuation, relaxing the Shapley axioms leads to weighting schemes for marginal utilities that improve the quality of values

pyDVL: the python Data Valuation Library

pyDVL strives to offer reference implementations of algorithms for data valuation and influence function computation, with an emphasis on …

Local calibration: metrics and recalibration

A calibration method that takes sample similarity into account, automatically providing group calibration even when groups are unknown.

Variational- and Metric-based Deep Latent Space for Out-of-Distribution Detection

A new method for detection of out-of-distribution samples in classification with neural networks which does not require OOD training data.

Improving cooperative game theory-based data valuation via data utility learning

Data valuation is typically limited by the exponential complexity of retraining for each possible subset of training points. This paper …

Boosting Certified Robustness of L Infinity Distance Nets

Recent architectural improvements for neural network classifiers make them certifiably robust against certain types of adversarial attacks - …

Data sheets for datasets

Gathering and documenting data is another process which requires attention for a successful application of ML. With standardized data …

Faster AutoAugment: Learning Augmentation Strategies using Backpropagation

Automatic augmentation is nowadays standard, but can be very slow. A new search strategy greatly speeds up the selection of augmentation …

Deep physical neural networks trained with backpropagation

Deep networks need not be artificial. In this paper the author build physical networks and train them in-silico. This can dramatically …

BarrierNet: A Safety-Guaranteed Layer for Neural Networks

A strategy for encoding constraints on the state space of a control problem as a differentiable layer for neural network based controllers …

Certified error rates for neural networks

Adversarial training has been shown to improve robustness of neural networks to certain classes of data perturbations. Despite constant …

Cross-validation: what does it estimate?

In Cross-validation: what does it estimate and how well does it do it? Bates et al. focus on cross validation for error estimation and show …

Data (e)valuation and model interpretation: a game theoretic approach

Attributing a “fair” (for some definition of fair) value to training samples has multiple applications. It can be used to investigate or …

Calibration of modern classifiers

An introduction to the pitfalls of uncalibrated classifiers and modern techniques for (re)calibration.

Solving PDEs With Neural Networks

We discuss some current trends in applying ML to the solution of differential equations, and the difficulties faced. We focus on a family of …

K.A.L.E. - The calibration game

The calibration game simulates resource allocation (medicine) for healthcare. A limited budget constraints prescription to patients with …

kyle - A toolkit for classifier calibration

There exist a number of different methods for calibrating models, and for measuring and visualizing calibration. Given the lack of reliable …

The calibrated classifier

Informed decision-making based on classifiers requires that the confidence in their predictions reflect the actual error rates. When this …

A short introduction to model-based Reinforcement Learning

Model-based RL learns a deterministic or probabilistic forward model of the dynamics of an environment for action planning or direct policy …

Explaining classifier outputs with local approximations

We learn about local perturbation methods for interpretation and review SLIME.

The case for learned index structures

Indexes are models: a B-Tree-Index can be seen as a model to map a key to the position of a record within a sorted array, a Hash-Index as a …