Getting Started with Deep Learning for Software Developers
summary
Shows a concrete Java path to deep learning: load a CSV credit card dataset (~284k rows with ~500 frauds), balance classes by sampling, train a feed‑forward neural network with backprop, and evaluate with confusion matrix metrics. Uses the Deep Nets open‑source Java API and a builder-style model/ trainer, reaching 97% train / 89% test on that example. Also demos a UI wizard, a Twitter engagement bot, and notes current limitations like no GPU support yet and work with Tornado VM to add it.
AI-written summary · about
chapters (10)
- 0:00Welcome and introduction
- 2:09Why deep learning matters now
- 15:05Hello world binary classification
- 22:04Data preparation essentials
- 27:12Building models in Java
- 35:21Standardizing ML APIs
- 40:48Real-world Twitter bot demo
- 44:27Java deep learning libraries
- 46:38Getting started resources
- 47:10Q&A on model evaluation
related talks
06mInterview
51mMachine Learning Exposed Deep and Reinforcement Learning
44mMachine Learning Exposed
50mApplying Machine Learning to IoT Use Cases with Deeplearning4j
44mDeep Learning with DeepLearning4J and Spring Boot
26mMachine Learning in the Browser with Deep Neural Networks
01h 01mDeep Learning and Java
35mDeconstructing Deep Learning