A5042985365- Data Mining Algorithms and Applications
- Insurance, Mortality, Demography, Risk Management
- Advanced Clustering Algorithms Research
- Natural Language Processing Techniques
- Gene expression and cancer classification
- Acute Ischemic Stroke Management
- Speech Recognition and Synthesis
- Text and Document Classification Technologies
- Artificial Intelligence in Healthcare
- Language Development and Disorders
- Nutritional Studies and Diet
Pontifícia Universidade Católica de Minas Gerais
2020-2025
Characterizing longevity profiles from longitudinal studies is a task with many challenges. Firstly, the databases usually have high dimensionality, and similarities between long-lived non-long-lived records are highly burdening for profile characterization. Addressing these issues, in this work, we use data English Longitudinal Study of Ageing (ELSA-UK) to characterize through mining. We propose method feature engineering reducing dimensionality merging techniques, factor analysis...
The understanding of the variables that influence human ageing process can help in development public policies improve quality life elderly population. In previous experiments on a dataset, we applied biclustering techniques binary data order to find long-lived and non-long lived profiles, but only found profiles. Then this work, propose use Factor Analysis represent reduced dimensionality, generating three datasets where with high correlation each other belong same factor. We observed some...
Human aging is a complex process with several factors interacting. One of the ways to identify patterns about human longitudinal population studies. In this work, we identified longevity profiles through knowledge discovery. After identifying profiles, apply triadic rules which allow extracting implication conditions. These can be used related factors, in various waves, studies, better explain conditions that favor profiles.The results show analysis efficient temporal evolution clinical or...
Failure to express yourself verbally is a condition that affects nearly 7% of children worldwide, known as specific language impairment. The diagnosis complex, involving specialists such speech therapists and pediatricians. dataset used in this work has many attributes imbalanced data, which can harm knowledge discovery. We biclustering identify clusters characterize with problems those typical development. propose selecting through the significance analysis biclusters, enhanced F-score...