Definition of Loss Functions for Learning from Imbalanced Data to Minimize Evaluation Metrics

2014 ◽  
pp. 19-37
Author(s):  
Juan Miguel García-Gómez ◽  
Salvador Tortajada
Keyword(s):  
Imbalanced Data ◽  
Loss Functions ◽  
Evaluation Metrics ◽  
Definition Of

Corporate failure prediction models in the twenty-first century: a review

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
David Veganzones ◽  
Eric Severin
Keyword(s):  
Prediction Models ◽  
Failure Prediction ◽  
First Century ◽  
Evaluation Metrics ◽  
Prediction Methods ◽  
Content Type ◽  
Corporate Failure ◽  
Twenty First Century ◽  
Definition Of ◽  
Corporate Failure Prediction

Purpose Corporate failure remains a critical financial concern, with implications for both firms and financial institutions; this paper aims to review the literature that proposes corporate failure prediction models for the twenty-first century. Design/methodology/approach This paper gathers information from 106 published articles that contain corporate failure prediction models. The focus of the analysis is on the elements needed to design corporate failure prediction models (definition of failure, sample approach, prediction methods, variables and evaluation metrics and performance). The in-depth review creates a synthesis of current trends, from the view of those elements. Findings Both consensus and divergences emerge regarding the design of corporate failure prediction models. On the one hand, authors agree about the use of bankruptcy as a definition of failure and that at least two evaluation metrics are needed to examine model performance for each class, individually and in general. On the other hand, they disagree about data collection procedures. Although several explanatory variables have been considered, all of them serve as complements for the primarily used financial information. Finally, the selection of prediction methods depends entirely on the research objective. These discrepancies suggest fundamental advances in discovery and establish valuable ideas for further research. Originality/value This paper reveals some caveats and provides extensive, comprehensible guidelines for corporate failure prediction, which researchers can leverage as they continue to investigate this critical financial subject. It also suggests fruitful directions to develop further experiments.

scholarly journals Feature selected cost-sensitive twin SVM for imbalanced data

2020 ◽  
Vol 309 ◽  
pp. 05013
Author(s):  
Xiaopeng Li ◽  
Xianrong Zhang
Keyword(s):  
Feature Selection ◽  
High Dimension ◽  
Imbalanced Data ◽  
Experimental Results ◽  
Evaluation Metrics ◽  
Cost Of Misclassification

In this paper, we propose a cost-sensitive twin SVM (cs-tsvm) and apply it to imbalanced data. A weight is added to each instance according to its cost of misclassification which is related to its position. In preprocessing part, features are selected by their difference of majority and minority classes. The feature is selected when its difference value is higher than average one. The experiment is conducted on UCI datasets and G-mean, AUC and accuracy are evaluation metrics. The experimental results show that Feature selection with CS-TWSVM is useful for datasets with high dimension.

scholarly journals Deep Attentive Ranking Networks for Learning to Order Sentences

2020 ◽  
Vol 34 (05) ◽  
pp. 8115-8122
Author(s):  
Pawan Kumar ◽  
Dhanajit Brahma ◽  
Harish Karnick ◽  
Piyush Rai
Keyword(s):  
Loss Function ◽  
State Of The Art ◽  
Loss Functions ◽  
Evaluation Metrics ◽  
Invariant Representation ◽  
Art Methods ◽  
Ranking Loss ◽  
Order Invariant

We present an attention-based ranking framework for learning to order sentences given a paragraph. Our framework is built on a bidirectional sentence encoder and a self-attention based transformer network to obtain an input order invariant representation of paragraphs. Moreover, it allows seamless training using a variety of ranking based loss functions, such as pointwise, pairwise, and listwise ranking. We apply our framework on two tasks: Sentence Ordering and Order Discrimination. Our framework outperforms various state-of-the-art methods on these tasks on a variety of evaluation metrics. We also show that it achieves better results when using pairwise and listwise ranking losses, rather than the pointwise ranking loss, which suggests that incorporating relative positions of two or more sentences in the loss function contributes to better learning.

scholarly journals A Comprehensive Survey on Image Dehazing Based on Deep Learning

2021 ◽  
Author(s):  
Jie Gui ◽  
Xiaofeng Cong ◽  
Yuan Cao ◽  
Wenqi Ren ◽  
Jun Zhang ◽  
...  
Keyword(s):  
Deep Learning ◽  
Loss Functions ◽  
Evaluation Metrics ◽  
Future Research ◽  
Image Dehazing ◽  
The Core ◽  
Comprehensive Survey ◽  
Open Issues

The presence of haze significantly reduces the quality of images. Researchers have designed a variety of algorithms for image dehazing (ID) to restore the quality of hazy images. However, there are few studies that summarize the deep learning (DL) based dehazing technologies. In this paper, we conduct a comprehensive survey on the recent proposed dehazing methods. Firstly, we conclude the commonly used datasets, loss functions and evaluation metrics. Secondly, we group the existing researches of ID into two major categories: supervised ID and unsupervised ID. The core ideas of various influential dehazing models are introduced. Finally, the open issues for future research on ID are pointed out.

scholarly journals Classification assessment methods

2020 ◽  
Vol ahead-of-print (ahead-of-print) ◽  
Author(s):  
Alaa Tharwat
Keyword(s):  
Confusion Matrix ◽  
Imbalanced Data ◽  
Classification Algorithms ◽  
Classification Problems ◽  
Operating Characteristics ◽  
Trade Off ◽  
Numerical Examples ◽  
Assessment Measures ◽  
Multi Class Classification ◽  
Definition Of

Classification techniques have been applied to many applications in various fields of sciences. There are several ways of evaluating classification algorithms. The analysis of such metrics and its significance must be interpreted correctly for evaluating different learning algorithms. Most of these measures are scalar metrics and some of them are graphical methods. This paper introduces a detailed overview of the classification assessment measures with the aim of providing the basics of these measures and to show how it works to serve as a comprehensive source for researchers who are interested in this field. This overview starts by highlighting the definition of the confusion matrix in binary and multi-class classification problems. Many classification measures are also explained in details, and the influence of balanced and imbalanced data on each metric is presented. An illustrative example is introduced to show (1) how to calculate these measures in binary and multi-class classification problems, and (2) the robustness of some measures against balanced and imbalanced data. Moreover, some graphical measures such as Receiver operating characteristics (ROC), Precision-Recall, and Detection error trade-off (DET) curves are presented with details. Additionally, in a step-by-step approach, different numerical examples are demonstrated to explain the preprocessing steps of plotting ROC, PR, and DET curves.

scholarly journals The Importance of Loss Functions for Increasing the Generalization Abilities of a Deep Learning-Based Next Frame Prediction Model for Traffic Scenes

2020 ◽  
Vol 2 (2) ◽  
pp. 78-98 ◽  
Author(s):  
Sandra Aigner ◽  
Marco Körner
Keyword(s):  
Deep Learning ◽  
Prediction Model ◽  
Short Term Memory ◽  
Mean Squared Error ◽  
Signal To Noise Ratio ◽  
Similarity Index ◽  
Structural Similarity ◽  
Loss Functions ◽  
Evaluation Metrics ◽  
Training Dataset

This paper analyzes in detail how different loss functions influence the generalization abilities of a deep learning-based next frame prediction model for traffic scenes. Our prediction model is a convolutional long-short term memory (ConvLSTM) network that generates the pixel values of the next frame after having observed the raw pixel values of a sequence of four past frames. We trained the model with 21 combinations of seven loss terms using the Cityscapes Sequences dataset and an identical hyper-parameter setting. The loss terms range from pixel-error based terms to adversarial terms. To assess the generalization abilities of the resulting models, we generated predictions up to 20 time-steps into the future for four datasets of increasing visual distance to the training dataset—KITTI Tracking, BDD100K, UA-DETRAC, and KIT AIS Vehicles. All predicted frames were evaluated quantitatively with both traditional pixel-based evaluation metrics, that is, mean squared error (MSE), peak signal-to-noise ratio (PSNR), and structural similarity index (SSIM), and recent, more advanced, feature-based evaluation metrics, that is, Fréchet inception distance (FID), and learned perceptual image patch similarity (LPIPS). The results show that solely by choosing a different combination of losses, we can boost the prediction performance on new datasets by up to 55%, and by up to 50% for long-term predictions.

Introductory paper

1966 ◽  
Vol 24 ◽  
pp. 3-5
Author(s):  
W. W. Morgan
Keyword(s):  
Line Intensity ◽  
Classification Scheme ◽  
Two Dimensional ◽  
Spectral Classification ◽  
Dimensional Array ◽  
Rr Lyrae ◽  
Metallic Line ◽  
Introductory Paper ◽  
Parameter Varying ◽  
Definition Of

1. The definition of “normal” stars in spectral classification changes with time; at the time of the publication of theYerkes Spectral Atlasthe term “normal” was applied to stars whose spectra could be fitted smoothly into a two-dimensional array. Thus, at that time, weak-lined spectra (RR Lyrae and HD 140283) would have been considered peculiar. At the present time we would tend to classify such spectra as “normal”—in a more complicated classification scheme which would have a parameter varying with metallic-line intensity within a specific spectral subdivision.

2.2. Working Group 2: Conceptual Definitions of Reference Coordinate Systems for Earth Dynamics

1975 ◽  
Vol 26 ◽  
pp. 21-26
Keyword(s):  
Coordinate System ◽  
Working Group ◽  
General Character ◽  
Coordinate Systems ◽  
Reference Coordinate System ◽  
Group 2 ◽  
Definition Of ◽  
Earth Dynamics

An ideal definition of a reference coordinate system should meet the following general requirements:1. It should be as conceptually simple as possible, so its philosophy is well understood by the users.2. It should imply as few physical assumptions as possible. Wherever they are necessary, such assumptions should be of a very general character and, in particular, they should not be dependent upon astronomical and geophysical detailed theories.3. It should suggest a materialization that is dynamically stable and is accessible to observations with the required accuracy.

Symbiotic Stars at Optical, Infrared and Radio Wavelengths

1979 ◽  
Vol 46 ◽  
pp. 125-149 ◽  
Author(s):  
David A. Allen
Keyword(s):  
Temperature Regime ◽  
Type Star ◽  
Stellar Systems ◽  
Symbiotic Stars ◽  
Lower Excitation ◽  
Definition Of ◽  
Original Type

No paper of this nature should begin without a definition of symbiotic stars. It was Paul Merrill who, borrowing on his botanical background, coined the termsymbioticto describe apparently single stellar systems which combine the TiO absorption of M giants (temperature regime ≲ 3500 K) with He II emission (temperature regime ≳ 100,000 K). He and Milton Humason had in 1932 first drawn attention to three such stars: AX Per, CI Cyg and RW Hya. At the conclusion of the Mount Wilson Ha emission survey nearly a dozen had been identified, and Z And had become their type star. The numbers slowly grew, as much because the definition widened to include lower-excitation specimens as because new examples of the original type were found. In 1970 Wackerling listed 30; this was the last compendium of symbiotic stars published.

Immunoferritin localization of intracellular antigens in positively stained frozen sections

1978 ◽  
Vol 36 (2) ◽  
pp. 168-169
Author(s):  
K. T. Tokuyasu
Keyword(s):  
Surface Tension ◽  
Water Surface ◽  
Thin Layers ◽  
The Other ◽  
Positive Staining ◽  
Frozen Sections ◽  
The Past ◽  
Ultrathin Frozen Sections ◽  
Definition Of

During the past investigations of immunoferritin localization of intracellular antigens in ultrathin frozen sections, we found that the degree of negative staining required to delineate u1trastructural details was often too dense for the recognition of ferritin particles. The quality of positive staining of ultrathin frozen sections, on the other hand, has generally been far inferior to that attainable in conventional plastic embedded sections, particularly in the definition of membranes. As we discussed before, a main cause of this difficulty seemed to be the vulnerability of frozen sections to the damaging effects of air-water surface tension at the time of drying of the sections.Indeed, we found that the quality of positive staining is greatly improved when positively stained frozen sections are protected against the effects of surface tension by embedding them in thin layers of mechanically stable materials at the time of drying (unpublished).

Sign in / Sign up

Export Citation Format

Share Document