"Systems Engineering Techniques for Visual Neuroscience", Dr. Alireza Ghahari

Dear all,
 

Neural Computation Unit (Doya Unit) would like to invite you to a seminar as follows.
--

Date: Thursday, February 8
Time: 11:00-
Venue: Seminar Room B503 - Ctr Bldg
 

Speaker: Dr. Alireza Ghahari, National Eye Institute, NIH​

Title: Systems Engineering Techniques for Visual Neuroscience

Abstract: Research in visual neuroscience addresses how neuronal population coding in vertebrate retina mediates the broad range of visual functions. Timely, elaborate research progress in this field is driven forward by iterative, exploratory cycles of development, technical validation, experiments, and performance optimization. This has provided unique opportunity for design creativity and exploration of diverse approaches to deliver services and software solutions with an acute sensitivity to the challenges of the neuroscience research spectrum. In particular, systems engineering techniques have posed as one of the potent and efficient ways, enabling much synergy and coherence in long-term strategic planning for research groups. Moreover, they have reduced the essential burden of systematic engineering an essential yet time-consuming part of the process-allowing neuroscientists to redirect their focus toward activities at the frontiers.

During this seminar, I will share the highlights of my career focusing on systems engineering techniques suited to visual neuroscience. In the first part, I explain how my PhD research culminated in neural systems and circuits design approaches to examine the role of the reticular formation of the pons and midbrain as the basic circuitry that mediates saccadic eye movements. The second part concerns my postdoctoral research in which the main thrust has been implementing integrated machine learning routines for classifying the signals that retinal ganglion cells in mouse model use to communicate with the brain. Both these studies have targeted the fundamental goal of estimating the code the language of electrical signals that the central nervous system uses to carry information to and from the eye. They have involved massive efforts on data mining and machine learning approaches to elucidate the sensorimotor control and its base of spiking neural networks in visual neuroscience. Consequently, these efforts have led to improving the quality and optimal use of neuromimetic computing algorithms in this technically demanding field. To facilitate the unprecedented scale of data mining tasks and gain improvement in both performance and speed, it is possible to generalize the data-driven and distributed core of these approaches into other neuroscientific domains.

--

We hope to see many of you. 
Sincerely, 
Emiko Asato
Neural Computation Unit