| |
|
|
| |
EVENT DETAILS |
Deep Learning is the biggest change happening in computer science right now. It powers everything from Google's Alpha Go to Tesla's autopilot to Amazon's Echo. Every company is trying to figure what it's AI strategy is going to be. Deep learning makes all kinds of new applications possible but presents a whole new set of challenges like exotic hardware & non-determinism. That's why companies can't hire experts fast enough. We strongly believe you don't need a degree from Stanford or MIT to build your own algorithms & use this amazing technology.
While there are plenty of online resources, we know it's tough to learn a technical topic without a teacher. We're bringing together expert engineers in the field of machine learning, deep learning & AI who will help you learn the basics in a hands-on approach to learning deep learning.
This is an extremely hands-on course to take students from little knowledge of deep learning to comfort building real world models. It requires very little math, but reasonably proficient programming skills. At the end of class students will be able to build models on their own, & more importantly be able to quickly find resources to help them with new problems they encounter in their domain.
Technologies Introduced
Deep Learning Frameworks
Keras
Tensorflow
Caffe
Torch
Deep Learning Model Architectures
Multi-Layer Perceptron
Convolutional Neural Network
LSTM
Applications
Object detection
Image segmentation
Bounding Boxes
What you need to bring:
Students need to bring a machine with python 3 installed on it along with the numpy library installed. If students bring a laptop with an nvidia GPU we will set the libraries up to take advantage of them. If students want to rent a GPU from EC2 or Google Cloud we can use that as well.
Take aways:
-Practical high-level knowledge of how deep learning algorithms actually work
-Familiarity with popular frameworks, TensorFlow, Caffe, Torch, especially Keras
-How to install the frameworks so they take advantage of your GPUs
-How to build models from scratch
-How to fine tune popular models like Inception & ResNet when training data is limited
Curriculum:
Day 1)
Morning: Introduction to Neural Nets
9:00 - 10:00 Breakfast & Laptop Setup
10:00 - 11:00 Refresher on machine learning, High-level overview of deep learning with & without calculus
11:00- 12:00 Build a small neural network from scratch together in python to do digit recognition
12:00-1:00 Lunch
Afternoon: Overview of Deep Learning Frameworks
1:00-2:00 Introduction to CUDA, cudNN & compiling programs to use GPU
2:00-3:00 Tensorflow introduction build a model in tensorflow to classify digits.
3:00-4:00 Caffe & Torch introduction. Build & export models to classify objects.
Day 2)
Morning: Training models in Keras
9:00-10:00 Building a real world model from scratch in Keras
10:00-12:00 Overview of Keras libraries.
Learn how to efficiently get data in & out
Make our image classifier do bounding boxes
Make our image classifier do pixel segmentation
Build a test classifier using word2vec & LSTM
Afternoon: Deep Learning in the Real World
1:00-3:00 How to make Keras/Tensorflow models really work in your job
-How to actually deploy models
-How to shrink model size with 8-bit quantization
-How to improve performance by fine tuning existing models
4:00-5:00 Where to go from here
-Resources for learning more about deep learning
Testimonials & Feedback
"I found it to be really engaging & interesting. I was already familiar with some ML concepts, so it helped me understand them better & think about how to apply them. The code samples are really great & will definitely reference them in the future. I thought the class went at a generally good pace."
"Good experience - full of great resources & discussion. Good, practical intro for new folks, & also valuable for those familiar with the basics. I walked away excited to experiment!"
Class was great, you ticked off my curiosity. I am excited to review the content & retry it by myself. Thank you for encouraging peer to peer collaboration & making the effort to build the slack channel. I think it was nice to see you debug live.
Corporate Training:
We also do corporatetrainings- send us an email for details.
|
|
|
|
|
|
