Uncertainty Estimation for Black-Box Classification Models: A Use Case for Sentiment Analysis

Mena, José; Brando, Axel; Pujol, Oriol; Vitrià, Jordi

doi:10.1007/978-3-030-31332-6_3

Cited by 8 publications

(7 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the simplest form, inheritance can be attained by sharing the same training data, i.e., by retraining the models [ 54 ]. Alternatively, model wrappers can also be used to envelope the existing solutions with an additional learnable layer that enables adaptation [ 55 , 56 ]. Other inheritance methods include data editing [ 57 , 58 , 59 ] or data enrichment.…”

Section: Adapting Models To the Demands Of Their Environmentmentioning

confidence: 99%

Differential Replication for Credit Scoring in Regulated Environments

Unceta

Nin

Pujol

2021

Entropy

Self Cite

View full text Add to dashboard Cite

Differential replication is a method to adapt existing machine learning solutions to the demands of highly regulated environments by reusing knowledge from one generation to the next. Copying is a technique that allows differential replication by projecting a given classifier onto a new hypothesis space, in circumstances where access to both the original solution and its training data is limited. The resulting model replicates the original decision behavior while displaying new features and characteristics. In this paper, we apply this approach to a use case in the context of credit scoring. We use a private residential mortgage default dataset. We show that differential replication through copying can be exploited to adapt a given solution to the changing demands of a constrained environment such as that of the financial market. In particular, we show how copying can be used to replicate the decision behavior not only of a model, but also of a full pipeline. As a result, we can ensure the decomposability of the attributes used to provide explanations for credit scoring models and reduce the time-to-market delivery of these solutions.

show abstract

Section: Adapting Models To the Demands Of Their Environmentmentioning

confidence: 99%

Differential Replication for Credit Scoring in Regulated Environments

Unceta

Nin

Pujol

2021

Entropy

Self Cite

View full text Add to dashboard Cite

show abstract

“…Hence, the re-training approach may not always be the most efficient nor the most effective way for tackling this challenge. Another option is adding patches in the form of wrappers to already deployed models to endow them with new traits or functionalities that help them adapt to the new data conditions, either globally [ 13 ] or locally [ 14 ]. Another example is that where a company wants to focus on a new client portfolio.…”

Section: Survival Of the Fittestmentioning

confidence: 99%

“…This form of inheritance requires no access to the parent model, but assumes knowledge of its training data. In addition to re-training, model wrappers can be used to envelope the existing solutions with an additional learnable layer that enables adaptation [ 13 , 24 ] (It is worth mentioning that the family of wrappers may require access to the model internals. In this study we classify them in this category by considering the most agnostic and general case).…”

Section: Differential Replicationmentioning

confidence: 99%

See 1 more Smart Citation

Environmental Adaptation and Differential Replication in Machine Learning

Unceta

Nin

Pujol

2020

Entropy

Self Cite

View full text Add to dashboard Cite

When deployed in the wild, machine learning models are usually confronted with an environment that imposes severe constraints. As this environment evolves, so do these constraints. As a result, the feasible set of solutions for the considered need is prone to change in time. We refer to this problem as that of environmental adaptation. In this paper, we formalize environmental adaptation and discuss how it differs from other problems in the literature. We propose solutions based on differential replication, a technique where the knowledge acquired by the deployed models is reused in specific ways to train more suitable future generations. We discuss different mechanisms to implement differential replications in practice, depending on the considered level of knowledge. Finally, we present seven examples where the problem of environmental adaptation can be solved through differential replication in real-life applications.

show abstract

“…In this work, we have a different reason for considering the original model as a black box. Similarly to [33,34] but applied to regression problems, the main goal here is not to generate a new model that imitates the previous one but to generate a new wrapper model that complements and maintains the original prediction with new information: the uncertainty score, as shown in Figure 1.…”

Section: Related Workmentioning

confidence: 99%

Building Uncertainty Models on Top of Black-Box Predictive APIs

et al. 2020

Self Cite

View full text Add to dashboard Cite

With the commoditization of machine learning, more and more off-the-shelf models are available as part of code libraries or cloud services. Typically, data scientists and other users apply these models as "black boxes" within larger projects. In the case of regressing a scalar quantity, such APIs typically offer a predict() function, which outputs the estimated target variable (often referred to aŝ y or, in code, y_hat). However, many real-world problems may require some sort of deviation interval or uncertainty score rather than a single point-wise estimate. In other words, a mechanism is needed with which to answer the question "How confident is the system about that prediction?" Motivated by the lack of this characteristic in most predictive APIs designed for regression purposes, we propose a method that adds an uncertainty score to every black-box prediction. Since the underlying model is not accessible, and therefore standard Bayesian approaches are not applicable, we adopt an empirical approach and fit an uncertainty model using a labelled dataset (x, y) and the outputsŷ of the black box. In order to be able to use any predictive system as a black box and adapt to its complex behaviours, we propose three variants of an uncertainty model based on deep networks. The first adds a heteroscedastic noise component to the black-box output, the second predicts the residuals of the black box, and the third performs quantile regression using deep networks. Experiments using real financial data that contain an inproduction black-box system and two public datasets (energy forecasting and biology responses) illustrate and quantify how uncertainty scores can be added to black-box outputs. INDEX TERMS Aleatoric uncertainty, Deep learning, Neural networks, Regression problems.

show abstract

Uncertainty Estimation for Black-Box Classification Models: A Use Case for Sentiment Analysis

Cited by 8 publications

References 7 publications

Differential Replication for Credit Scoring in Regulated Environments

Differential Replication for Credit Scoring in Regulated Environments

Environmental Adaptation and Differential Replication in Machine Learning

Building Uncertainty Models on Top of Black-Box Predictive APIs

Contact Info

Product

Resources

About