Events
Artificial Intelligence Workshop 2021
Nov 15-21
Event Concluded
Keynotes
Session I. Artificial Intelligence Demystified
Session II. How to learn Data Science?
Session III. Succeeding in a Data Science Career
Workshops
WORKSHOP I. Deep Learning for Beginners
Session 1 (2 hours): Foundation of Deep Learning
- Summary of Machine Learning to segue to Deep Learning
- Understanding a problem and its data
- Formulating a solution
- Loss function
- Model fitting
- Develop "Hello World" Deep Learning Model
- Deconstruct and understand the operations in Multilayer Perceptron
- Backpropagation Method
- Hands-on development
- Q&A
Session 2 (2 hours): Advanced models in Deep Learning
- LSTM Models
- Fundamentals of LSTM
- Memory cells
- Information flow
- Hands-on modeling
- CNN Models
- Fundamentals of CNN
- Filtration process
- Properties that make CNN standout
- Hands-on modeling
- Q&A
WORKSHOP II. Computer Vision and Deep Learning
Session 1 (2 Hours): Image Data and Machine Learning
- Image data fundamentals
- Image processing
- Manipulation using OpenCV Python
- Data augmentation using Pytorch Python
- Machine learning introduction
- Hands on live-coding
- Q&A
Session 2 (2 hours): Deep Learning for Images
- Deep Learning Introduction
- Deep Learning for Images using Pytorch
- Image Classification
- Image Segmentation
- State of the art overview
- GAN models using images
- Practical applications
- Hands on live-coding
- Practical modeling and application tips and tricks
- Q&A
WORKSHOP III. ML-OPS: Deep Learning Development on AWS Sagemaker
• Background of Cloud Computing in AWS (Amazon Web Services)
• Introduction to Machine Learning Development in Amazon SageMaker
• Create your own AWS Account for continuous hands-on exercise
• Python Jupyter Notebooks
- show and create examples of simple typical notebooks
• Modeling in SageMaker
- Hands-on with real-data
- Introduction to special features and algorithms
- Data Preprocessing and Feature Engineering using Amazon SageMaker Data Wrangler
- Import from various data sources (add examples)
- Data exploration and visualization
- Cleansing (outliers, missing values)
- Normalize; standardize; transform; scale; time synchronization
- Estimate model accuracy
• Hyper Parameter Optimization: Amazon SageMaker Automatic Model Training and Tuning
- Automatically finding the best model
- Three different ways of doing
• Amazon SageMaker Studio: IDE for Machine Learning
- Automatically build and train Models: Amazon SageMaker AutoPilot
- Learn the IDE for ML to automatically transform data, build, train, debug, deploy, and monitor your ML models
• Amazon SageMaker Experiments: Organize and Keep Track of Modeling Results
- Organize, track, compare, and evaluate your ML experiments
- Observe the impact of incremental changes on model accuracy
Speakers
Chitta Ranjan, Ph.D., Director of Science at ProcessMiner | Author of, "Understanding Deep Learning."
Subhajit Das, Ph.D., Applied Scientist, Amazon Inc.
Esa Vilkama, Chief of Data Science/ML at CloudTrust Inc. | Founder of Process Data Insights.