Deep learning is a technique for automatically finding hierarchical patterns in large sets of data, and has been used to achieve breakthrough advances in computer vision, machine translation, speech recognition, game playing, robotics, and other applications in recent years. The recent progress and future potential of deep learning has led to immense interest and to its adoption by all large technology companies.
In this workshop, we’ll introduce deep learning and demonstrate how it can be used in the above areas, with a focus on practical applications. We will explain the technological and algorithmic advances that have made it possible, describe the tools you can use to get started, and talk about the challenges to deploying deep learning systems in production.
What you will learn
In this workshop, you will learn the core techniques used in Deep Learning. Through examples in Keras, TensorFlow and Apache Spark, you will learn
- Basics of Neural Networks
- Core Deep Learning Techniques including CNNs, RNNs, AutoEncoders
- Limitations and challenges of using deep neural networks.
- Fine tuning and considerations when working with Deep Neural Networks
- Deep Learning and Apache Spark
- Practical case studies with fully functional code
After this workshop, you will be able to:
- Describe what deep neural networks (DNNs) are, what they can be used for, and how they fit with other AI techniques.
- Explain what computer games have to do with the success of deep learning.
- Describe several deep learning technologies and their tradeoffs.
- Explain the limitations and challenges of using DNNs.
- Train and tune simple DNNs and evaluate their performance, using Python and several deep learning frameworks.
- Describe how DNNs are likely to affect your field in the next 5-10 years.
- Work on case studies in Keras,TensorFlow and Apache Spark
In module one, we will review the core techniques in Deep learning neural networks. Through examples we will understand the different deep learning techniques and frameworks
- Introduction to deep neural networks
- Hands on with Keras and TensorFlow
- Convolutional neural networks
- Recurrent neural networks for translation, sentiment detection, and other text applications
- Case study 1: Classifying images using fully connected neural networks and convolutional networks
- Case study 2: Monitoring network learning and status using ad-hoc plots and TensorBoard
- Case study 3: Comparing performance of GPU and CPU network training
- Case study 4: Using pre-trained models for identifying objects in photos.
In module two, we will discuss more advanced techniques in deep learning. We will also discuss best practices in scaling and using deep learning techniques including using Apache Spark with Deep learning.
- Deep neural nets for reinforcement learning--games, robots, and self-driving cars
- Hands on: tuning parameters, initial values, optimizers, and other practical issues
- Recurrent Neural Networks, AutoEncoders and Reinforcement Learning
- Apache Spark and Deep Learning
- Intro to Microsoft Cognitive Toolkit
- Frontiers of deep learning: what’s coming in the next few years
- Case study 5: Understanding the word2vec text embedding
- Case study 6: Using recurrent neural networks to caption images.
- Case study 7: Using an auto-encoder to de-noise images.
- Case study 8: Using neural-network-based reinforcement learning to play breakout
Sample Content from a prior workshop
Past Attendees of QuantUniversity workshops include Assette, Baruch College, Bentley College, Bloomberg, BNY Mellon, Boston University, Datacamp, Fidelity, Ford, Goldman Sachs, IBM, J.P. Morgan Chase, MathWorks, Matrix IFS, MIT Lincoln Labs, Morgan Stanley, Nataxis Global, Northeastern University, NYU, Pan Agora, Philips Health, Stevens Institute, T.D. Securities and many more..