update README

Author: Edyarich

README.rst

@@ -15,26 +15,21 @@
 
 .. class:: center
 
-    :Название исследуемой задачи: Мое название работы
-    :Тип научной работы: M1P/НИР/CoIS
-    :Автор: Имя Отчество Фамилия
-    :Научный руководитель: степень, Фамилия Имя Отчество
-    :Научный консультант(при наличии): степень, Фамилия Имя Отчество
+    :Название исследуемой задачи: Причинно-ориентированное снижение размерности для анализа данных нейроинтерфейсов
+    :Тип научной работы: НИР
+    :Автор: Владимиров Эдуард Анатольевич
+    :Научный руководитель: д.ф-м.н. Стрижов Вадим Викторович
+    :Научный консультант(при наличии): -
 
 Abstract
 ========
 
-Purus sit amet luctus venenatis lectus magna fringilla. Condimentum id venenatis a condimentum vitae sapien. Auctor neque vitae tempus quam pellentesque nec nam aliquam. Eget egestas purus viverra accumsan in. Nam aliquam sem et tortor. Donec et odio pellentesque diam volutpat commodo. Odio pellentesque diam volutpat commodo sed egestas. Sit amet dictum sit amet justo. Pellentesque sit amet porttitor eget. Imperdiet sed euismod nisi porta lorem mollis aliquam ut. Quis varius quam quisque id diam vel quam elementum. Ornare lectus sit amet est placerat in. Vulputate dignissim suspendisse in est ante. Elit sed vulputate mi sit amet mauris. Egestas sed tempus urna et pharetra pharetra. Viverra tellus in hac habitasse platea dictumst vestibulum rhoncus.
+Learning low–dimensional representations that preserve cause–and–effect structure is a key step toward interpretable modelling of high-dimensional dynamical data.
+		This thesis introduces \textbf{CaSCA}, a linear auto-encoder that splits the latent space into a causal block, capturing delayed directed influence from one multivariate series to another, and a reconstructive block, keeping residual variance.  
+		Extensions to trajectory embeddings, Riemannian covariance spaces, and a deep variant with a differentiable Convergent Cross Mapping loss broaden the framework.  
+		Comprehensive experiments on two real-world datasets—dual accelerometer-gyroscope recordings and EEG–IMU traces of table-tennis sessions—show that CaSCA  
+		(i) reduces multicollinearity,  
+		(ii) reconstructs signals with negligible loss of explained variance, and  
+		(iii) improves downstream prediction
+		The method thus offers a compact, interpretable state-space where causal links are easier to detect and exploit.
 
-Research publications
-===============================
-1. 
-
-Presentations at conferences on the topic of research
-================================================
-1. 
-
-Software modules developed as part of the study
-======================================================
-1. A python package *mylib* with all implementation `here <https://github.com/intsystems/ProjectTemplate/tree/master/src>`_.
-2. A code with all experiment visualisation `here <https://github.comintsystems/ProjectTemplate/blob/master/code/main.ipynb>`_. Can use `colab <http://colab.research.google.com/github/intsystems/ProjectTemplate/blob/master/code/main.ipynb>`_.