-
Understanding challenges to the interpretation of disaggregated evaluations of algorithmic fairness
Authors:
Stephen R. Pfohl,
Natalie Harris,
Chirag Nagpal,
David Madras,
Vishwali Mhasawade,
Olawale Salaudeen,
Awa Dieng,
Shannon Sequeira,
Santiago Arciniegas,
Lillian Sung,
Nnamdi Ezeanochie,
Heather Cole-Lewis,
Katherine Heller,
Sanmi Koyejo,
Alexander D'Amour
Abstract:
Disaggregated evaluation across subgroups is critical for assessing the fairness of machine learning models, but its uncritical use can mislead practitioners. We show that equal performance across subgroups is an unreliable measure of fairness when data are representative of the relevant populations but reflective of real-world disparities. Furthermore, when data are not representative due to sele…
▽ More
Disaggregated evaluation across subgroups is critical for assessing the fairness of machine learning models, but its uncritical use can mislead practitioners. We show that equal performance across subgroups is an unreliable measure of fairness when data are representative of the relevant populations but reflective of real-world disparities. Furthermore, when data are not representative due to selection bias, both disaggregated evaluation and alternative approaches based on conditional independence testing may be invalid without explicit assumptions regarding the bias mechanism. We use causal graphical models to predict metric stability across subgroups under different data generating processes. Our framework suggests complementing disaggregated evaluations with explicit causal assumptions and analysis to control for confounding and distribution shift, including conditional independence testing and weighted performance estimation. These findings have broad implications for how practitioners design and interpret model assessments given the ubiquity of disaggregated evaluation.
△ Less
Submitted 4 June, 2025;
originally announced June 2025.
-
Microbial assessment in a rare Norwegian book collection: a One Health approach to cultural heritage
Authors:
Sílvia O. Sequeira,
Ekaterina Pasnak,
Carla Viegas,
Bianca Gomes,
Marta Dias,
Renata Cervantes,
Pedro Pena,
Magdalena Twarużek,
Robert Kosicki,
Susana Viegas,
Liliana Aranha Caetano,
Maria João Penetra,
Inês Santos,
Ana Teresa Caldeira,
Catarina Pinheiro
Abstract:
Microbial contamination poses a threat to both the preservation of library and archival collections and the health of staff and users. This study investigated the microbial communities and potential health risks associated with the UNESCO-classified Norwegian Sea Trade Archive (NSTA) collection exhibiting visible microbial colonization and staff health concerns. Dust samples from book surfaces and…
▽ More
Microbial contamination poses a threat to both the preservation of library and archival collections and the health of staff and users. This study investigated the microbial communities and potential health risks associated with the UNESCO-classified Norwegian Sea Trade Archive (NSTA) collection exhibiting visible microbial colonization and staff health concerns. Dust samples from book surfaces and the storage environment were analysed using culturing methods, qPCR, Next Generation Sequencing, and mycotoxin, cytotoxicity and azole resistance assays. Penicillium sp., Aspergillus sp., and Cladosporium sp. were the most common fungi identified, with some potentially toxic species like Stachybotrys sp., Toxicladosporium sp. and Aspergillus section Fumigati. Fungal resistance to azoles was not detected. Only one mycotoxin, sterigmatocystin, was found in a heavily contaminated book. Dust extracts from books exhibited moderate to high cytotoxicity on human lung cells, suggesting a potential respiratory risk. The collection had higher contamination levels compared to the storage environment, likely due to improved storage conditions. Even though, overall low contamination levels were obtained, which might be underestimated due to the presence of salt (from cod preservation) that could have interfered with the analyses. This study underlines the importance of monitoring microbial communities and implementing proper storage measures to safeguard cultural heritage and staff well-being.
△ Less
Submitted 29 March, 2024;
originally announced April 2024.
-
Representation of Compact Operators between Banach spaces
Authors:
G. Ramesh,
M. Veena Sangeetha,
Shanola S. Sequeira
Abstract:
In this article, we give a representation for compact operators acting between reflexive Banach spaces, which generalizes the representation given by Edmunds et al. for compact operators between reflexive Banach spaces with strictly convex duals. Further, we give a representation for operators on Banach spaces that are comparable to compact normal operators on Hilbert spaces and illustrate our res…
▽ More
In this article, we give a representation for compact operators acting between reflexive Banach spaces, which generalizes the representation given by Edmunds et al. for compact operators between reflexive Banach spaces with strictly convex duals. Further, we give a representation for operators on Banach spaces that are comparable to compact normal operators on Hilbert spaces and illustrate our result with an example.
△ Less
Submitted 15 August, 2023;
originally announced August 2023.
-
Conditions implying the normality of $\ast$-paranormal operators in the closure of $\mathcal{AN}$-operators
Authors:
G. Ramesh,
Shanola S. Sequeira
Abstract:
In this article, we first prove the existence of an invariant subspace for a norm-attaining $\ast$-paranormal operator. Then give a representation for $\ast$-paranormal operators in the closure of absolutely norm-attaining operators and further study a few sufficient conditions for the normality of such operators. Finally, we discuss Toeplitz and Hankel $\ast$-paranormal operators in the closure o…
▽ More
In this article, we first prove the existence of an invariant subspace for a norm-attaining $\ast$-paranormal operator. Then give a representation for $\ast$-paranormal operators in the closure of absolutely norm-attaining operators and further study a few sufficient conditions for the normality of such operators. Finally, we discuss Toeplitz and Hankel $\ast$-paranormal operators in the closure of absolutely norm-attaining operators on the Hardy space.
△ Less
Submitted 2 February, 2023;
originally announced February 2023.
-
Absolutely minimum attaining Toeplitz and absolutely norm attaining Hankel operators
Authors:
G. Ramesh,
Shanola S. Sequeira
Abstract:
In this article, we completely characterize absolutely norm attaining Hankel operators and absolutely minimum attaining Toeplitz operators. We also improve \cite[Theorem 2.1]{RGSSSTOE1}, by characterizing the absolutely norm attaining Toeplitz operator $T_\varphi$ in terms of the symbol $\varphi \in L^\infty$.
In this article, we completely characterize absolutely norm attaining Hankel operators and absolutely minimum attaining Toeplitz operators. We also improve \cite[Theorem 2.1]{RGSSSTOE1}, by characterizing the absolutely norm attaining Toeplitz operator $T_\varphi$ in terms of the symbol $\varphi \in L^\infty$.
△ Less
Submitted 13 September, 2022;
originally announced September 2022.
-
Representation and normality of Hyponormal operators in the closure of $\mathcal{AN}$-operators
Authors:
G. Ramesh,
Shanola S. Sequeira
Abstract:
Let $H_1$, $H_2$ be complex Hilbert spaces. A bounded linear operator $T : H_1 \to H_2$ is said to be norm attaining if there exists a unit vector $x \in H_1$ such that $\|Tx\| = \|T\|$. If $T|_{M} : M \to H_2$ is norm attaining for every closed subspace $M$ of $H_1$, then we say that $T$ is an absolutely norm attaining ($\mathcal{AN}$-operator). If the norm of the operator is replaced by the mini…
▽ More
Let $H_1$, $H_2$ be complex Hilbert spaces. A bounded linear operator $T : H_1 \to H_2$ is said to be norm attaining if there exists a unit vector $x \in H_1$ such that $\|Tx\| = \|T\|$. If $T|_{M} : M \to H_2$ is norm attaining for every closed subspace $M$ of $H_1$, then we say that $T$ is an absolutely norm attaining ($\mathcal{AN}$-operator). If the norm of the operator is replaced by the minimum modulus $m(T) = \inf\{\|Tx\| : x \in H_1, \|x\| =1\}$, then $T$ is said to be a minimum attaining and an absolutely minimum attaining operator ($\mathcal{AM}$-operator), respectively.
In this article, we give representations of quasinormal $\mathcal{AN}$, $\mathcal{AM}$-operators and the operators in the closure of these two classes. Later we extend these results to the class of hyponormal operators in the closure of $\mathcal{AN}$-operators and a further look at some sufficient conditions under which these operators become normal.
△ Less
Submitted 13 August, 2022;
originally announced August 2022.
-
On the closure of Absolutely Norm attaining Operators
Authors:
G. Ramesh,
Shanola S. Sequeira
Abstract:
Let $H_1$ and $H_2$ be complex Hilbert spaces and $T:H_1\rightarrow H_2$ be a bounded linear operator. We say $T$ to be norm attaining, if there exists $x\in H_1$ with $\|x\|=1$ such that $\|Tx\|=\|T\|$. If for every closed subspace $M$ of $H_1$, the restriction $T|_{M}:M\rightarrow H_2$ is norm attaining then, $T$ is called absolutely norm attaining operator or $\mathcal{AN}$-operator. If we repl…
▽ More
Let $H_1$ and $H_2$ be complex Hilbert spaces and $T:H_1\rightarrow H_2$ be a bounded linear operator. We say $T$ to be norm attaining, if there exists $x\in H_1$ with $\|x\|=1$ such that $\|Tx\|=\|T\|$. If for every closed subspace $M$ of $H_1$, the restriction $T|_{M}:M\rightarrow H_2$ is norm attaining then, $T$ is called absolutely norm attaining operator or $\mathcal{AN}$-operator. If we replace the norm of the operator by the minimum modulus $m(T)=\inf{\{\|Tx\|:x\in H_1,\; \|x\|=1}\}$, then $T$ is called the minimum attaining and the absolutely minimum attaining operator (or $\mathcal{AM}$-operator) respectively.
In this article, we discuss about the operator norm closure of the $\mathcal{AN}$-operators. We completely characterize operators in this closure and study several important properties. We mainly give the spectral characterization of the positive operators in this class and give the representation when the operator is normal. Later we also study the analogous properties for $\mathcal{AM}$-operators and prove that the closure of $\mathcal{AM}$-operators is same as that of the closure of $\mathcal{AN}$-operators. As a consequence, we prove similar results for operators in the norm closure of $\mathcal{AM}$-operators.
△ Less
Submitted 12 April, 2022;
originally announced April 2022.
-
Underspecification Presents Challenges for Credibility in Modern Machine Learning
Authors:
Alexander D'Amour,
Katherine Heller,
Dan Moldovan,
Ben Adlam,
Babak Alipanahi,
Alex Beutel,
Christina Chen,
Jonathan Deaton,
Jacob Eisenstein,
Matthew D. Hoffman,
Farhad Hormozdiari,
Neil Houlsby,
Shaobo Hou,
Ghassen Jerfel,
Alan Karthikesalingam,
Mario Lucic,
Yian Ma,
Cory McLean,
Diana Mincu,
Akinori Mitani,
Andrea Montanari,
Zachary Nado,
Vivek Natarajan,
Christopher Nielson,
Thomas F. Osborne
, et al. (15 additional authors not shown)
Abstract:
ML models often exhibit unexpectedly poor behavior when they are deployed in real-world domains. We identify underspecification as a key reason for these failures. An ML pipeline is underspecified when it can return many predictors with equivalently strong held-out performance in the training domain. Underspecification is common in modern ML pipelines, such as those based on deep learning. Predict…
▽ More
ML models often exhibit unexpectedly poor behavior when they are deployed in real-world domains. We identify underspecification as a key reason for these failures. An ML pipeline is underspecified when it can return many predictors with equivalently strong held-out performance in the training domain. Underspecification is common in modern ML pipelines, such as those based on deep learning. Predictors returned by underspecified pipelines are often treated as equivalent based on their training domain performance, but we show here that such predictors can behave very differently in deployment domains. This ambiguity can lead to instability and poor model behavior in practice, and is a distinct failure mode from previously identified issues arising from structural mismatch between training and deployment domains. We show that this problem appears in a wide variety of practical ML pipelines, using examples from computer vision, medical imaging, natural language processing, clinical risk prediction based on electronic health records, and medical genomics. Our results show the need to explicitly account for underspecification in modeling pipelines that are intended for real-world deployment in any domain.
△ Less
Submitted 24 November, 2020; v1 submitted 6 November, 2020;
originally announced November 2020.
-
A low-cost real-time 3D imaging system for contactless asthma observation
Authors:
Sheona M. M. D. P. Sequeira,
Beril Sirmacek
Abstract:
Asthma is becoming a very serious problem with every passing day, especially in children. However, it is very difficult to detect this disorder in them, since the breathing motion of children tends to change when they reach an age of 6. This, thus makes it very difficult to monitor their respiratory state easily. In this paper, we present a cheap non-contact alternative to the current methods that…
▽ More
Asthma is becoming a very serious problem with every passing day, especially in children. However, it is very difficult to detect this disorder in them, since the breathing motion of children tends to change when they reach an age of 6. This, thus makes it very difficult to monitor their respiratory state easily. In this paper, we present a cheap non-contact alternative to the current methods that are available. This is using a stereo camera, that captures a video of the patient breathing at a frame rate of 30Hz. For further processing, the captured video has to be rectified and converted into a point cloud. The obtained point clouds need to be aligned in order to have the output with respect to a common plane. They are then converted into a surface mesh. The depth is further estimated by subtracting every point cloud from the reference point cloud (the first frame). The output data, however, when plotted with respect to real time produces a very noisy plot. This is filtered by determining the signal frequency by taking the Fast Fourier Transform of the breathing signal. The system was tested under 4 different breathing conditions: deep, shallow and normal breathing and while coughing. On its success, it was tested with mixed breathing (combination of normal and shallow breathing) and was lastly compared with the output of the expensive 3dMD system. The comparison showed that using the stereo camera, we can reach to similar sensitivity for respiratory motion observation. The experimental results show that, the proposed method provides a major step towards development of low-cost home-based observation systems for asthma patients and care-givers.
△ Less
Submitted 3 November, 2019;
originally announced November 2019.
-
Deployment characterization of a floatable tidal energy converter on a tidal channel, Ria Formosa, Portugal
Authors:
A. Pacheco,
E. Gorbeña,
T. A. Plomaritis,
E. Garel,
J. M. S. Gonçalves,
L. Bentes,
P. Monteiro,
C. M. L. Afonso,
F. Oliveira,
C. Soares,
F. Zabel,
S. Sequeira
Abstract:
This paper presents the results of a pilot experiment with an existing tidal energy converter (TEC), Evopod 1 kW floatable prototype, in a real test case scenario (Faro Channel, Ria Formosa, Portugal). A baseline marine geophysical, hydrodynamic and ecological study based on the experience collected on the test site is presented. The collected data was used to validate a hydro-morphodynamic model,…
▽ More
This paper presents the results of a pilot experiment with an existing tidal energy converter (TEC), Evopod 1 kW floatable prototype, in a real test case scenario (Faro Channel, Ria Formosa, Portugal). A baseline marine geophysical, hydrodynamic and ecological study based on the experience collected on the test site is presented. The collected data was used to validate a hydro-morphodynamic model, allowing the selection of the installation area based on both operational and environmental constraints. Operational results related to the description of power generation capacity, energy capture area and proportion of energy flux are presented and discussed, including the failures occurring during the experimental setup. The data is now available to the scientific community and to TEC industry developers, enhancing the operational knowledge of TEC technology concerning efficiency, environmental effects, and interactions (i.e. device/environment). The results can be used by developers on the licensing process, on overcoming the commercial deployment barriers, on offering extra assurance and confidence to investors, who traditionally have seen environmental concerns as a barrier, and on providing the foundations whereupon similar deployment areas can be considered around the world for marine tidal energy extraction.
△ Less
Submitted 25 September, 2019;
originally announced September 2019.
-
Transitions to Intermittency and Collective Behavior in Randomly Coupled Map Networks
Authors:
D. Volchenkov,
S. Sequeira,
Ph. Blanchard
Abstract:
We study the transitions to spatio-temporal intermittency in networks of randomly coupled Chate-Manneville maps. The relevant paprameters are the network connectivity, coupling strength, and the local parameter of the map. We show that the spatio-temporal intermittency occurs for some intervals or windows of the values of these parameters. Within the intermittency windows, the system exhibits pe…
▽ More
We study the transitions to spatio-temporal intermittency in networks of randomly coupled Chate-Manneville maps. The relevant paprameters are the network connectivity, coupling strength, and the local parameter of the map. We show that the spatio-temporal intermittency occurs for some intervals or windows of the values of these parameters. Within the intermittency windows, the system exhibits periodic and other nontrivial collective behaviors. The detailed behavior depends crucially upon the topology of the random graph spanning the network. We present a detailed analysis of the results based on the thermodynamic formalism and random graph theory.
△ Less
Submitted 17 May, 2001; v1 submitted 10 April, 2001;
originally announced April 2001.
