Intro to Deep Learning - ML Tech Talks

Dive into a comprehensive 1-hour 14-minute ML Tech Talk that provides an in-depth overview of Deep Learning. Explore representation learning, various neural network families and their applications, and gain insights into the inner workings of deep neural networks. Learn through numerous code examples and key concepts from TensorFlow. Discover the differences between AI, ML, and DL, understand hyperparameters, and explore the skills crucial for machine learning. Examine real-world applications in healthcare, investigate different neural network architectures, and delve into topics like overfitting, underfitting, and autoencoders. Benefit from book recommendations and access a wealth of helpful resources, including demos, tutorials, and additional courses to further your deep learning journey.

- Intro and outline
- TensorFlow.js demos + discussion
- AI vs ML vs DL
- What’s representation learning?
- A cartoon neural network more on this later
- What features does a network see?
- The “deep” in “deep learning”
- Why tree-based models are still important
- How your workflow changes with DL
- A couple illustrative code examples
- What’s a hyperparameter?
- The skills that are important in ML
- An example of applied work in healthcare
- Families of neural networks + applications
- Encoder-decoders + more on representation learning
- Families of neural networks continued
- Are neural networks opaque?
- Building up from a neuron to a neural network
- A demo of representation learning in TF Playground
- Importance of activation functions
- What’s a neural network library?
- Overfitting and underfitting
- Autoencoders and anomaly detection screencast and demo
- Book recommendations