Breaking the curse of dimensionality: hierarchical Bayesian network model for multi-view clustering

Njah, Hasna; Jamoussi, Salma; Mahdi, Walid

doi:10.1007/s10472-021-09749-z

Cited by 3 publications

(2 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This paper uses AEs as a dimensionality reduction technique to study their impact on six clustering techniques: AGGLO, BIRCH, KNN-EUC, KNN-DTW, MNBT, and SPCT 1 . The data utilized in this research is intraday data from IBEX, CAC, DAX, SPX, and UKX.…”

Section: Proposalmentioning

confidence: 99%

See 1 more Smart Citation

Autoencoder-Enhanced Clustering: A Dimensionality Reduction Approach to Financial Time Series

Cortés,

Onieva,

López

et al. 2024

IEEE Access

View full text Add to dashboard Cite

While Machine Learning significantly boosts the performance of predictive models, its efficacy varies across different data dimensions. It is essential to cluster time series data of similar characteristics, particularly in the financial sector. However, clustering financial time series data poses considerable challenges due to the market's inherent complexity and multidimensionality. To address these issues, our study introduces a novel clustering framework that leverages autoencoders for a compressed yet informative representation of financial time series. We rigorously evaluate our approach through multiple dimensionality reduction and clustering algorithms, applying it to key financial indices, including IBEX-35, CAC-40, DAX-30, S&P 500, and FTSE 100. Our findings consistently demonstrate that incorporating autoencoders significantly enhances the granularity and quality of clustering, effectively isolating distinct categories of financial time series. Our findings carry significant ramifications for the financial industry. By refining clustering methodologies, we set the stage for increasingly accurate financial predictive models, offering valuable insights for optimizing investment strategies and enhancing risk management.

show abstract

Section: Proposalmentioning

confidence: 99%

“…In addition, price information derived from financial markets at any given time is often presented as a series of opening, high, low, and Closing Prices (CP) and transaction volume. This multidimensional nature hinders efficient examination of the data and impedes the proper use of clustering techniques [1].…”

Section: Introductionmentioning

confidence: 99%

Autoencoder-Enhanced Clustering: A Dimensionality Reduction Approach to Financial Time Series

Cortés,

Onieva,

López

et al. 2024

IEEE Access

View full text Add to dashboard Cite

show abstract

Unpacking Occupational Health Data in the Service Sector: From Bayesian Networking and Spatial Clustering to Policy-Making

Pazo,

Boente,

Albuquerque

et al. 2023

Math Geosci

View full text Add to dashboard Cite

The health status of the service sector workforce is a significant unknown in the field of medical geography. While spatial epidemiology has made progress in predicting the relationship between human health and the environment, there are still important challenges that remain unsolved. The main issue lies in the inability to statistically determine and visually represent all spatial concepts, as there is a need to cover a wide range of service activities while also considering the impact of numerous traditional medical variables and emerging risk factors, such as those related to socioeconomic and bioclimatic factors. This study aims to address the needs of health professionals by defining, prioritizing, and visualizing multiple occupational health risk factors that contribute to the well-being of workers. To achieve this, a methodological approach based on the synergy of Bayesian machine learning and geostatistics is proposed. Extensive data from occupational health surveillance tests were collected in Spain, along with socioeconomic and bioclimatic covariates, to assess potential social and climate impacts on health. This integrated approach enabled the identification of relevant patterns related to risk factors. A three-step geostatistical modeling process, including variography, ordinary kriging, and G clustering, was used to generate national distribution maps for various factors such as annual mean temperature, annual rainfall, spine health, limb health, cholesterol, age, and sleep quality. These maps considered four target activities—administration, finances, education, and hospitality. Remarkably, bioclimatic variables were found to contribute approximately 9% to the overall health status of workers.

show abstract

Is it possible to find the single nearest neighbor of a query in high dimensions?

Ting,

Washio,

Zhu

et al. 2024

Artificial Intelligence

View full text Add to dashboard Cite

Breaking the curse of dimensionality: hierarchical Bayesian network model for multi-view clustering

Cited by 3 publications

References 21 publications

Autoencoder-Enhanced Clustering: A Dimensionality Reduction Approach to Financial Time Series

Autoencoder-Enhanced Clustering: A Dimensionality Reduction Approach to Financial Time Series

Unpacking Occupational Health Data in the Service Sector: From Bayesian Networking and Spatial Clustering to Policy-Making

Is it possible to find the single nearest neighbor of a query in high dimensions?

Contact Info

Product

Resources

About