Search

Search

Deep Learning Using TensorFlow

Deep Learning Using TensorFlow

Main Speaker:


Eran Kaufman

Tracks:

Code
Data

Seminar Catgories:

AI & Machine Learning
Back-end

Seminar ID:

42189

Date:

18.11.2019

Time:

Daily seminar
9:00-16:30

Location:

Daniel Hotel

Add to Calendar 18.11.2019 09:00 18.11.2019 16:30 Asia/Jerusalem Deep Learning Using TensorFlow

Recommendation system are system that help users to discover new products and services. Every time you shop online the system learns how to guide you using your ratings of other products.

Expert systems are system that you help get a better diagnosis and solution to a problem. These have been widely deployed in medicine, industry etc.

Vision systems help you recognize what’s in a photo, whether to help you build an album on Facebook or in surveillance cameras around the world to help track suspects.

In 2015 Google announced that all it’s engines (including the search engine and YouTubes recommendation system) will work on Neural Nets using the python package TensorFlow.

They have been backing up the package as an open source ever since.

In today’s world, no matter what your interest is in, machine learning is an important tool to have. This course is a programer level course to deep learning. Giving practical insight and hands-on tools for decision making and assessment of the tools and algorithms you may choose.

Overview

Recommendation system are system that help users to discover new products and services. Every time you shop online the system learns how to guide you using your ratings of other products.

Expert systems are system that you help get a better diagnosis and solution to a problem. These have been widely deployed in medicine, industry etc.

Vision systems help you recognize what’s in a photo, whether to help you build an album on Facebook or in surveillance cameras around the world to help track suspects.

In 2015 Google announced that all it’s engines (including the search engine and YouTubes recommendation system) will work on Neural Nets using the python package TensorFlow.

They have been backing up the package as an open source ever since.

In today’s world, no matter what your interest is in, machine learning is an important tool to have. This course is a programer level course to deep learning. Giving practical insight and hands-on tools for decision making and assessment of the tools and algorithms you may choose.

Who Should Attend

This course is intended for algorithm engineer, programers or business analysts who are taking their first steps with deep learning in order to provides them the required skills for becoming a productive data scientist in that environment.

Prerequisites

Participant must have:

  • engineering/scientific/mathematical academic background.
  • Experience working with python.

Course Contents

what is machine learning –  Basic concepts

  • Supervised learning: classification vs. regression.
  • Unsupervised Learning.
  • Structured learning : Bayesian networks, Markovian random fields.
  • Reinforcement Learning.

Development cycle

  • Get data.
  • prepare data
  • Train Model.
  • Validate

Learning theory

  • Overfitting vs. under fitting in practice.
  • Learning curves and their power to answer practical question.
  • Bias vs. variance decomposition.
  • Regularization.

Introduction to Deep Learning and Neural Nets

  • Introduction to neuroscience and NN historical background
  • Introduction to TensorFlow native API and keras
  • Introduction To CNNs and Architectures
  • Introduction to RNNs and Architectures

Practical examples:

  • Convolutional NN – the MNIST challenge
  • Recurrent NN  – Shakespeare synthesis
  • Speech Recognition   –  CNN + RNN

 



Contact