MLTrain 3

Mar 03 - Apr 03, 2017

Children’s Healthcare of Atlanta Pediatric Technology Center, Atlanta

Ismion Inc. is pleased to announce our first MLTrain event for this year in Atlanta. Nick Vasiloglou and Alex Dimakis will cover several Machine Learning and TensorFlow topics. We have prepared a 2 day curriculum. You can register for each day individually or for both days.

Schedule

  • 3/3/2017
  • 3/4/2017

3/3/2017

-

Introduction to TensorFlow and Keras

This session is intended for beginners. The only requirements are:

  • Be familiar with python programming
  • Be able to install tensorFlow before the class date
  • Be familiar with basic Machine Learning Principles

After the completion of the session you will know the basic functionality of TensorFlow. You will be able to build simple models and also use it in data science projects.

-

Introduction to TensorFlow

  • MLTrain Introduction
  • Tensors Basics
  • Computational Graph Model
  • Graph Inspection & Visualization with TensorBoard
  • Basic Ops
-

TensorFlow Cool Features

  • Types, Constants and Placeholders
  • Variables (creation, initialization, mutation, saving/restoring)
  • Advanced Operations
  • Automatic Differentiation
  • Device/Hardware Management
  • Kinds of Parallelism and Distributed Computing (Synchronous vs Asynchronous)
  • Debugging
-

Optimization In TensorFlow

  • Objective Function
  • Gradients Computation
  • The tf.Optimizer Class
  • Predefined Optimizers (FtrlOptimizer, GradientDescent, Adagrad, SDCAOptimizer)
  • TF Linear Regression Model In 3 Lines
  • Predefined Losses
-

Overview of the tf.contrib.learn Package

  • The Estimator Class
  • Feature Columns & Feature Engineering
  • Input Processing
  • Linear Estimators
  • Logging and Debugging
  • Deep Models in tf.contrib.learn Package
  • Training Deep Models in tf.contrib.learn
  • Monitoring and Debugging with TensorBoard
  • Reading Data in TF
  • Canonical Data Format (Example: proto)
  • Input Readers
  • Higher Level interfaces with Keras

3/4/2017

-

Advanced Deep Learning

In this session you will learn how to use TensorFlow for building deep learning models for different application domains. The session emphasizes understanding models, how to use them and when to trust them.

In order to attend this session you are expected:

  • To have basic knowledge of TensorFlow. You can do that by going through the tutorials in the www.tensorflow.org
  • To be proficient in python
  • To have tensorFlow already installed on your machine
  • To have some data science prior experience or exposure
-

Linear Algebra

  • Sparse/Dense Matrix/Vectors Operations
  • Kronecker Products in TF
  • From Matrices to Tensors
  • Tensor Tiling: The Map Operator of TF
  • Reductions on Tensors
  • Limitations of TF
-

Deep Learning Architectures

  • Recurrent Neural Networks
  • Convolutional Networks
  • Generative Adversarial Networks
  • Word2Vec
  • Transfer Learning
-

Working with Images

  • Understanding and using Generative Models
  • Classifying images
  • Working with Text
  • Word2Vec
  • LSTMs for parsing Text
  • Text classification
  • Logic
  • Encoding logical rules
  • Simple AI tasks

Location

Children’s Healthcare of Atlanta Pediatric Technology Center 950 Atlantic Drive NW, Atlanta, GA 30332, United States