Exploring Neural Network Hidden Layer Activity Using Vector Fields

Cantareira, Gabriel Dias; Etemad, Elham; Paulovich, Fernando V.

doi:10.3390/info11090426

Cited by 13 publications

(12 citation statements)

References 27 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the past few years, the use of visualization tools and techniques gives a better insight into the propagation of data in NN hidden layers [32] exploring different aspects of NN training, topology, and parametrization [33]. Special attention is paid to the process of activation of neurons in hidden layers through sets of activation functions and data propagation within hidden layers of the network, and the results show that different activation results between hidden layers are essential in creating an efficient internal network architecture [34].…”

Section: Neural Network Model Descriptionmentioning

confidence: 99%

Development of Models for Children—Pedestrian Crossing Speed at Signalized Crosswalks

et al. 2021

View full text Add to dashboard Cite

Modeling the behavior of pedestrians is an important tool in the analysis of their behavior and consequently ensuring the safety of pedestrian traffic. Children pedestrians show specific traffic behavior which is related to cognitive development, and the parameters that affect their traffic behavior are very different. The aim of this paper is to develop a model of the children-pedestrian’s speed at a signalized pedestrian crosswalk. For the same set of data collected in the city of Osijek—Croatia, two models were developed based on neural network and multiple linear regression. In both cases the models are based on 300 data of measured children speed at signalized pedestrian crosswalks on primary city roads located near a primary school. As parameters, both models include the selected traffic infrastructure features and children’s characteristics and their movements. The models are validated on data collected on the same type of pedestrian crosswalks, using the same methodology in two other urban environments—the city of Rijeka, Croatia and Enna in Italy. It was shown that the neural network model, developed for Osijek, can be applied with sufficient reliability to the other two cities, while the multiple linear regression model is applicable with relatively satisfactory reliability only in Rijeka. A comparative analysis of the statistical indicators of reliability of these two models showed that better results are achieved by the neural network model.

show abstract

Section: Neural Network Model Descriptionmentioning

confidence: 99%

Development of Models for Children—Pedestrian Crossing Speed at Signalized Crosswalks

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The t-SNE and other similar techniques are robust against norm concentration, a typical characteristic of the curse of dimensionality [7], because of the shift-invariant similarities [69]. Although attaining good results if compared to other methods, in t-SNE projections, the observed distance between clusters is unreliable and sensitive to parametrization (perplexity parameter) [70], which can generate a misleading effect of cluster distance and shape resulting from the local nature of optimization and how hyperparameters are set up [8]. Nonetheless, t-SNE presents computational complexity equal to O(n 2 ), which impairs practical applications, but Maaten [67] developed an accelerated version that reaches O(n log n) through approximations.…”

Section: Stochastic Neighbor Embeddingmentioning

confidence: 99%

“…Furthermore, UMAP presents results as good as t-SNE (currently one of the widely used projection techniques) in visualization quality; nonetheless, it better preserves global data structures and is faster than t-SNE. In addition, UMAP was developed with machine learning applications in mind [8].…”

Section: Uniform Manifold Approximation For Dimension Reductionmentioning

confidence: 99%

“…In this context, a class of InfoVis techniques has been a stand out for some time: the multidimensional projections. These techniques generate injective mappings that transform multidimensional spaces into visual spaces (2D or 3D), preserving structures and similarities of the original spaces as much as possible, such as relationships between data instances or clusters' presence [6][7][8]. This transformation is accomplished through mathematical operations to embed the number of attributes to 2 or 3, allowing m-dimensional data objects to be represented in Cartesian space.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Getting over High-Dimensionality: How Multidimensional Projection Methods Can Assist Data Science

2022

View full text Add to dashboard Cite

The exploration and analysis of multidimensional data can be pretty complex tasks, requiring sophisticated tools able to transform large amounts of data bearing multiple parameters into helpful information. Multidimensional projection techniques figure as powerful tools for transforming multidimensional data into visual information according to similarity features. Integrating this class of methods into a framework devoted to data sciences can contribute to generating more expressive means of visual analytics. Although the Principal Component Analysis (PCA) is a well-known method in this context, it is not the only one, and, sometimes, its abilities and limitations are not adequately discussed or taken into consideration by users. Therefore, knowing in-depth multidimensional projection techniques, their strengths, and the possible distortions they can create is of significant importance for researchers developing knowledge-discovery systems. This research presents a comprehensive overview of current state-of-the-art multidimensional projection techniques and shows example codes in Python and R languages, all available on the internet. The survey segment discusses the different types of techniques applied to multidimensional projection tasks from their background, application processes, capabilities, and limitations, opening the internal processes of the methods and demystifying their concepts. We also illustrate two problems, from a genetic experiment (supervised) and text mining (non-supervised), presenting solutions through multidimensional projection application. Finally, we brought elements that reverberate the competitiveness of multidimensional projection techniques towards high-dimension data visualization, commonly needed in data sciences solutions.

show abstract

“…They use a 2D projection to explain instance activation using t-SNE and place instances with similar activation values together. Cantareira et al [41] introduce an approach to exploring a neural network's hidden layer activities and simplifying the inner working mechanism of such complex networks. Their novel technique focuses on comparing projections derived from multiple stages in a neural network and visualizing the differences in perception.…”

Section: Visualization For Interpreting Modelsmentioning

confidence: 99%

Random Forest Similarity Maps: A Scalable Visual Representation for Global and Local Interpretation

2021

Self Cite

View full text Add to dashboard Cite

Machine Learning prediction algorithms have made significant contributions in today’s world, leading to increased usage in various domains. However, as ML algorithms surge, the need for transparent and interpretable models becomes essential. Visual representations have shown to be instrumental in addressing such an issue, allowing users to grasp models’ inner workings. Despite their popularity, visualization techniques still present visual scalability limitations, mainly when applied to analyze popular and complex models, such as Random Forests (RF). In this work, we propose Random Forest Similarity Map (RFMap), a scalable interactive visual analytics tool designed to analyze RF ensemble models. RFMap focuses on explaining the inner working mechanism of models through different views describing individual data instance predictions, providing an overview of the entire forest of trees, and highlighting instance input feature values. The interactive nature of RFMap allows users to visually interpret model errors and decisions, establishing the necessary confidence and user trust in RF models and improving performance.

show abstract

Exploring Neural Network Hidden Layer Activity Using Vector Fields

Cited by 13 publications

References 27 publications

Development of Models for Children—Pedestrian Crossing Speed at Signalized Crosswalks

Development of Models for Children—Pedestrian Crossing Speed at Signalized Crosswalks

Getting over High-Dimensionality: How Multidimensional Projection Methods Can Assist Data Science

Random Forest Similarity Maps: A Scalable Visual Representation for Global and Local Interpretation

Contact Info

Product

Resources

About