10-18, 10:10–10:40 (Europe/Lisbon), Auditorium
ime series analysis is vital for trends and forecasting in finance and healthcare. This talk explores ARIMA models for clustering disease data in 48 European countries, using Python libraries like pmdarima and TimeSeriesSplit. We’ll examine how distance measures like Piccolo, Maharaj, and LPC reveal health disparities, showcasing how Python enhances public health surveillance and interventions.
Time series analysis is essential for understanding trends, seasonality, and forecasting across various domains like finance and healthcare. In public health, ARIMA models are crucial for clustering disease trend data. This talk showcases an analysis of health indicators across 48 European countries, emphasizing the role of Python libraries such as pmdarima for ARIMA modeling and TimeSeriesSplit for robust cross-validation. Also, explore the impact of distance measures—like Piccolo, Maharaj, and LPC—on clustering outcomes, uncovering significant geographic health disparities. This aplication, llustrates how Python tools can enhance public health surveillance and inform effective intervention strategies.
This talk is strutured as:
1. Description of the problem
2. Description of the approach
3. Results presentation
4. Discussion and conclusions
Beginner
What are the main topics of your talk? –Data Science with Python; Time series clustering
Alexandra Oliveira is a Data Scientist consultant, an integrated researcher at the Artificial Intelligence and Computer Sciences Laboratory (LIACC) of the University of Porto, and an Adjunct Professor at the School of Health of the Polytechnic of Porto. She holds a Ph.D. in Applied Mathematics from University of Porto, as well as two post-doctoral degrees in Artificial Intelligence.
For over 10 years at the School of Health of the Polytechnic Institute of Porto, she has been a professor in the areas of Biomathematics, Biostatistics, and Bioinformatics, teaching and supervising master's theses on various applications of data science/artificial intelligence in health.
Her doctoral thesis focused on Public Health applications, where she worked with data from the national HIV/AIDS notification system, modeling them considering all human and technical factors affecting these systems.
At LIACC, where she has been collaborating since 2018, she primarily supervises master's theses in computer engineering involving data analysis, mostly with applications in individual or public health.
With over 10 years of experience in data science, she has developed solutions in various fields such as individual and public health, well-being assessment, and risk management applied to food inspections. She has worked on creating questionnaires, data analysis, forecasting, interpretable predictive and classification models, developing suitable methodologies for extracting valuable knowledge from databases, and has experience in producing technical and scientific documentation.