Misclassification Cost

Misclassification cost focuses on quantifying and minimizing the consequences of incorrect predictions in classification tasks, aiming to optimize models beyond simple accuracy metrics. Current research explores methods like cost-sensitive boosting and adaptive learning algorithms, often applied within neural network architectures, to dynamically adjust model behavior based on the relative costs of different errors. This research is crucial for improving the reliability and robustness of machine learning systems in high-stakes applications where the cost of misclassification varies significantly, such as medical diagnosis or autonomous driving. The development of effective techniques for handling misclassification costs is essential for building trustworthy and impactful AI systems.

Papers

July 10, 2024

The Misclassification Likelihood Matrix: Some Classes Are More Likely To Be Misclassified Than Others
Daniel Sikar, Artur Garcez, Robin Bloomfield, Tillman Weyde, Kaleem Peeroo, Naman Singh, Maeve Hutchinson, Dany Laksono, Mirela Reljan-Delaney
Convolutional Neural Network MNIST Dataset Visual Based Class Decision Threshold Misclassification Cost

May 17, 2024

Submodular Information Selection for Hypothesis Testing with Misclassification Penalties
Jayanth Bhargav, Mahsa Ghasemi, Shreyas Sundaram
Hypothesis Testing Information Gain Greedy Algorithm Subset Selection Submodular Mutual Information Optimal Set Misclassification Cost Bayesian Setting

January 4, 2023

Cost-Sensitive Stacking: an Empirical Evaluation
Natalie Lawrance, Marie-Anne Guerry, George Petrides
Real World Empirical Evaluation Cost Sensitive Stacking Ensemble Stacking Framework Misclassification Cost

September 17, 2022

AdaCC: Cumulative Cost-Sensitive Boosting for Imbalanced Classification
Vasileios Iosifidis, Symeon Papadopoulos, Bodo Rosenhahn, Eirini Ntoutsi
Class Imbalance Imbalanced Classification Cost Sensitive Misclassification Cost

July 19, 2022

Adaptive Learning for the Resource-Constrained Classification Problem
Danit Shifman Abukasis, Izack Cohen, Xiaochen Xian, Kejun Huang, Gonen Singer
Adaptive Learning Classification Problem Misclassification Cost

May 11, 2022

Analysis of convolutional neural network image classifiers in a rotationally symmetric model
Michael Kohler, Benjamin Walter
Convolutional Neural Network General Analysis Symmetric Model Misclassification Cost Binary Image Classification

January 31, 2022

Agnostic Learnability of Halfspaces via Logistic Loss
Ziwei Ji, Kwangjun Ahn, Pranjal Awasthi, Satyen Kale, Stefani Karp
Logistics System Agnostic Learning Margin Halfspaces Approximation Guarantee Misclassification Cost

November 14, 2021

Adaptive Cost-Sensitive Learning in Neural Networks for Misclassification Cost Problems
Ohad Volk, Gonen Singer
Neural Network Cost Sensitive Sensitivity Aware Distribution Mismatch Misclassification Cost