Difference between revisions of "My fist scientific paper"
| Line 65: | Line 65: | ||
==Functional Data Analysis, 2024== | ==Functional Data Analysis, 2024== | ||
| − | The statistical analysis of spatial time series requires additional methods of data analysis. First, we suppose time is continuous, put to the state space changes <math>\frac{d\mathbf{x}}{dt}</math> and use neural ordinary and stochastic differential equations. Second, we analyze a multivariate and multidimensional time series and use the tensor representation and tensor analysis. Third, since the time series have significant cross-correlation we model them in the Riemannian space. Fourth, medical time series are periodic, the base model is the pendulum model, <math>\frac{d^2x}{dt^2}=-c\sin{x}</math>. We use physics-informed neural networks to approximate data. Fifth, the practical experiments involve multiple data sources. We use canonical correlation analysis with latent state space. This space aligns the source and target spaces and generates data in source and target manifolds. | + | The statistical analysis of spatial time series requires additional methods of data analysis. First, we suppose time is continuous, put to the state space changes <math>\frac{d\mathbf{x}}{dt}</math> and use neural ordinary and stochastic differential equations. Second, we analyze a multivariate and multidimensional time series and use the tensor representation and tensor analysis. Third, since the time series have significant cross-correlation we model them in the Riemannian space. Fourth, medical time series are periodic, the base model is the pendulum model, <math>\frac{d^2x}{dt^2}=-c\sin{x}</math>. We use physics-informed neural networks to approximate data. Fifth, the practical experiments involve multiple data sources. We use canonical correlation analysis with latent state space. This space aligns the source and target spaces and generates data in source and target manifolds. [[Functional Data Analysis|See the course page]]. |
<!--*[http://bit.ly/m1p_file2discuss Upload a file to discussion] | <!--*[http://bit.ly/m1p_file2discuss Upload a file to discussion] | ||
Revision as of 22:44, 15 September 2024
Fall 2024 on September 14 — The Art of Scientific Research
Fall 2024 on September 13 — Intelligent Data Analysis / FDA
Before January 2025 — My fist scientific paper: Suggest your project here
Spring 2025 on February 6th — My fist scientific paper starts
See results of 2024 — on GitHub
Contents
The Art of Scientific Research
See you this Saturday (September 21, 2024) at 14:40 m1p.org/go_zoom
The goal is to prepare and select the research topic of your dream. We must be sure that the problem statement and project planning lead you to successful delivery.
m1p Course progress
This course produces student research papers. It gathers research teams. Each team joins a student, a consultant, and an expert. The student is a project driver who wants to plunge into scientific research activities. The graduate student consultant conducts the research and helps the student. The expert, a professor, states the problem and enlightens the road to the goal. The projects start in February and end in May, according to the schedule.
- Course schedule, Spring 2025
- Week 0: Come in
- Week 1: Set the toolbox
- Week 2: Select your project and tell about it
- Week 3: State your problem
- Week 4: Plan the experiment
- Week 5: Visualise the principle
- Week 6: Write the theory
- Week 7: Analyse the error
- Week 8: Construct your paper
- Week 9: Review a paper
- Week 10: Select a journal to submit
- Week 11: Prepare your presentation
- Week 12: Show your results (Youtube)
LINKS
- 2024 results GitHub
- 2024 problems GitHub
- 2023 problems GitHub
- 2022 results GitHub
- Telegram: discussion Ask here!
- The meeting room: m1p.org/go_zoom
- More courses from the Intelligent Systems
HISTORY
- Main page with old homework
- Group 674, 694, spring 2020
- Group 674, spring 2019
- Group 694, spring 2019
Mathematical methods of forecasting, 2024
This course delivers methods of model selection in machine learning and forecasting. The modeling data are videos, audio, encephalograms, fMRIs, and other measurements in natural science. The models are linear, tensor, deep neural networks, and neural ODEs. The practical examples are brain-computer interfaces, weather forecasting, and various spatial-time series forecasting. The lab works are organized as paper-with-code reports. See the course page
Functional Data Analysis, 2024
The statistical analysis of spatial time series requires additional methods of data analysis. First, we suppose time is continuous, put to the state space changes \(\frac{d\mathbf{x}}{dt}\) and use neural ordinary and stochastic differential equations. Second, we analyze a multivariate and multidimensional time series and use the tensor representation and tensor analysis. Third, since the time series have significant cross-correlation we model them in the Riemannian space. Fourth, medical time series are periodic, the base model is the pendulum model, \(\frac{d^2x}{dt^2}=-c\sin{x}\). We use physics-informed neural networks to approximate data. Fifth, the practical experiments involve multiple data sources. We use canonical correlation analysis with latent state space. This space aligns the source and target spaces and generates data in source and target manifolds. See the course page.