A novel deep learning approach to predict crash severity in adverse weather on rural mountainous freeway

Khan, Nasim; Ahmed, Mohamed M.

doi:10.1080/19439962.2022.2129891

Cited by 7 publications

(2 citation statements)

References 50 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, Neto et al 47 illustrated that a convolutional layer with DeepInsight could match the performance of XGBoost on Arboviruses data. Similarly, Khan et al 48 13 , applied DeepInsight to the sparse MNIST database, and Pasquadibisceglie et al 15 , incorporated it within their ORANGE method for outcome predictions on diverse event logs. Tajmirriahi et al 17 , also drew inspiration from DeepInsight for P300 detection in timeseries EEG signal analysis.…”

Section: Discussionmentioning

confidence: 99%

Enhanced analysis of tabular data through Multi-representation DeepInsight

Sharma,

López,

Jia

et al. 2024

Sci Rep

View full text Add to dashboard Cite

Tabular data analysis is a critical task in various domains, enabling us to uncover valuable insights from structured datasets. While traditional machine learning methods can be used for feature engineering and dimensionality reduction, they often struggle to capture the intricate relationships and dependencies within real-world datasets. In this paper, we present Multi-representation DeepInsight (MRep-DeepInsight), a novel extension of the DeepInsight method designed to enhance the analysis of tabular data. By generating multiple representations of samples using diverse feature extraction techniques, our approach is able to capture a broader range of features and reveal deeper insights. We demonstrate the effectiveness of MRep-DeepInsight on single-cell datasets, Alzheimer's data, and artificial data, showcasing an improved accuracy over the original DeepInsight approach and machine learning methods like random forest, XGBoost, LightGBM, FT-Transformer and L2-regularized logistic regression. Our results highlight the value of incorporating multiple representations for robust and accurate tabular data analysis. By leveraging the power of diverse representations, MRep-DeepInsight offers a promising new avenue for advancing decision-making and scientific discovery across a wide range of fields.

show abstract

Section: Discussionmentioning

confidence: 99%

Enhanced analysis of tabular data through Multi-representation DeepInsight

Sharma,

López,

Jia

et al. 2024

Sci Rep

View full text Add to dashboard Cite

show abstract

“…RBFNet outperforms existing models, showing high accuracy in detecting respiratory diseases while accounting for skewed data distributions related to gender, age, and smoking status, underscoring its potential as a robust diagnostic tool. [31] [32] During machine learning training, the process of minimising empirical loss might unintentionally induce bias due to the presence of discrimination and social biases in the data. In order to overcome the limitations of conventional fair machine learning approaches, which often depend on sensitive information from training data or need substantial modifications to the model, we introduce FairIF, a distinctive two-stage training framework.…”

Section: Literature Reviewmentioning

confidence: 99%

Detection of Gender in Crowds Using ResNet Model

Rajeev G. Vishwakarma

2024

jes

View full text Add to dashboard Cite

The ResNet model is used in this investigation to suggest a gender detection solution for use in congested settings. Due to occlusions, varied stances, and various features, determining a person's gender in crowded surroundings may be a difficult and time-consuming job. The ResNet model, which is a deep convolutional neural network architecture, is used to solve these difficulties because of its capacity to capture detailed characteristics and its efficiency in managing deep network structures. The strategy that has been suggested entails preprocessing the input photos, sending those images through the ResNet model, and then extracting gender-related characteristics from those images. The ResNet model is made up of a number of residual blocks with skip connections, which makes it easier to learn complicated representations. After that, the learnt characteristics are input into fully linked layers, and then softmax activation is used to determine the subject's gender. The usefulness of the technique that was developed was shown by experimental findings on a large dataset, which achieved a high level of accuracy in gender determination. The use of the ResNet model helps the system to handle complicated scenarios and improves the system's ability to accurately recognize gender in situations with a large number of people. The method that has been developed has the potential to find applications in areas such as surveillance, crowd control, and the study of social behavior.

show abstract