Differences

This shows you the differences between two versions of the page.

--- hackathon_2015 [2015/10/27 11:01] – created n.arakawa
+++ hackathon_2015 [2016/01/27 11:00] (current) – n.arakawa
@@ Line 1: / Line 1: @@
-The first WBA hackathon was held in September 2015 organized by the Future Leaders of WBA and WBAI.
+The first WBA hackathon was held in September 2015, organized by [[http://wbawakate.jp/|the Future Leaders of WBA]] and WBAI.
   * Date: 2015-09-19...23
   * Venue: Hiyoshi Campus, Keio University (Yokohama, Japan)
-  * Participants: five teams (23 hackers) and 12 mentors
+  * Participants: seven teams (23 hackers) and 12 mentors
-  * Result: to be announced (on GitHub)
+  * Purpose: developing know-how and skills on using multiple machine learning modules
+  * Result: visit [[https://github.com/wbap/Hackathon2015|the GitHub site]] ([[http://wbawakate.jp/posts/events/第1回wbaiハッカソン活動報告/|report in Japanese]])
+  * Financial aids: domestic transportation & lodging
+  * Prizes:
+     * 1st: BICA2015 participation incl. transportation & lodging
+     * other prizes donated by sponsoring companies
 ----
 **The Ideathon**\\
 An ideathon was held in June to develop ideas on tasks for the hackathon.\\
-The criteria for ideas are as follows:
+The hackathon teams were selected based on their ideathon proposals.
+  * Ideas wanted: to combine machine learning modules referring to the brain to attain advanced functionality
+  * Contents to be proposed
+     * Functionality to be realized
+     * The way to realize the functionality
+     * Combination of machine learning modules
+     * Relevance to brain architecture
+     * Required resources (human & machine)
+  * Criteria for adopting
+     * Importance of the functionality
+     * Feasibility in the hackathon
+     * Originality/Room for further development
+     * Biological reality
+  * Sample ideas
+     * Video game play combining deep learning (visual cortex) and reinforcement learning (basal ganglia) (as in Deep-Q Network)
+     * Caption generation combining deep learning (visual cortex) and RNN (prefrontal/language cortex) (as in NeuralTalk @ Stanford U.)