A Raspberry Pi and DeepLearning4J Brain That Plays Minecraft
summary
Builds a Raspberry Pi cluster mapped to neocortex layers (one Pi 3 controller plus four Pi Zero Ws), runs Zulu JVM and DeepLearning4J (ND4J, RL4J) with Microsoft’s project Malmo mod for Minecraft, and implements a Markov decision / deep Q learning setup. Goal shifts from building a house to a simpler stay alive task: agent learns to avoid lava blocks from world-state vectors; demo partially works—agent sometimes survives but still often dies. Raspberry Pi sales exceed 10 million, showing cheap embedded hardware is viable for small-scale RL experiments.
AI-written summary · about
chapters (20)
- 0:00Intro
- 1:21Inspiration
- 3:09Human Brain Neocortex
- 4:30Hierarchical Temporal Memory
- 6:07Introduction To Machine Learning
- 14:01Neural Network Elements: Node
- 15:18Neural Network Elements: Layers
- 17:28Markov Decision Process
- 20:44Deep Q Network
- 24:05Raspberry Pi Variations
- 24:56Pi Brain Architecture
- 25:42Raspberry Pi Cluster (Mk 2)
- 26:53Cluster Hat
- 29:24Azul Zulu Java For Embedded
- 33:29RL4J Basics
- 38:31Malmo Example Code (2)
- 40:02Minecraft Interaction
- 40:50Demo Software Architecture
- 41:59Demo Scenario
- 44:07Conclusions
related talks
50mA Raspberry Pi and DeepLearning4J Brain That Plays Minecraft
01h 09mMinecraft as AI Playground and Laboratory
41mReinforcement Learning and Minecraft
50mIEEE CoG 2019 - Industry Day - Katja Hofmann (Microsoft Research)
28mAI and Minecraft Lars
32mAI and Minecraft
05mInterview with
50mApplying Machine Learning to IoT Use Cases with Deeplearning4j