Design and Deployment of Artificial Neural Networks with Spark
October 27 @ 8:00 am - 6:00 pm
This is a one-day, 8 hour bootcamp during which you will cover artificial neural network introduction to getting to production with optimized deployment.
TICKETS $130-195, SIGN UP THROUGH EVENTBRITE
( Door Prize of AI Frontier Conference 11/9-11/2018, $1,295 Business pass – All access pass, including evening keynote, dinner banquet with wine and socializing opportunities with speakers and other attendees. Detail on https://aifrontiers.com/ It is a showcase for new AI applications, AI startups, training day of Tensorflow from Google Brain. Sign up today for 30% off discount with code *p30ACMbay* at http://bit.ly/airegister )
Tickets: Single $150 from 9/13 to 10/15 at 3pm. After 10/15 at 3pm, the price for any ticket is $195. Before 10/15, group tickets 2-6 at a time, the price is $130 / person.
NO RSVP on MEETUP, TICKET PURCHASE on EVENTBRIGHT:
We are seeking TA’s who know ML to help the audience. TA applicants should contact the instructor in advance. Use the [contact] button on the left, send email, phone, LinkedIn and ML experience).
PARKING AND ENTRANCE:
Both at MLSListings Inc. in the BACK OF BUILDING.
PRE-LOADING: BEFORE THE CLASS, PREPARATIONS:
For all workshops we will use Jupiter notebooks with Python, Spark and BigDL.
Notebook instances will be provided by the organizers.
8 HR CLASS – SUMMARY (detailed outline follows) We will review the design of convolutional neural networks (CNNs) and their history. Why convolutional networks are designed in this particular way; what is the purpose of convolutions, pooling layers and fully connected layers; how CNNs relate to feed-forward (fully-connected) networks. In the workshop section we will apply this knowledge to create a LeNet-5 convolutional network using Spark/BigDL, train the network and use it to recognize handwritten digits from the MNIST dataset.
After that we will review the concept of transfer learning — a sophisticated method of applying large convolutional neural networks to real-world problems of recognizing arbitrary images. In the workshop sections we will use BigDL and pre-trained Inception Model and apply transfer learning concepts to recognize images from an arbitrary dataset.
You can expect to take-away from the workshop
.Theoretical underpinning of Deep Learning technologies
. Practical application of DL frameworks to business problems
Advanced track to come:
.Generative Adversarial Networks;
.Trainers available 1 week after-class office-hour to review your project.
TARGET AUDIENCE would include people who …
• are comfortable in programming
• may already work on consulting projects or in some technical business problem solving role.
• It is helpful if you have tried Python, Spark and BigDL before.
COURSE DESIGN: For students to get the most out of a one day class, the instructors are focusing on a “narrow” path, like a project sprint, going through a complete set of steps in a machine learning project. Many pointers will be provided to invite you to broaden your skills after the class.
The instructors like the Covey quote “If the ladder is not leaning against the right wall, every step we take just gets us to the wrong place faster.” A successful machine learning project is not just coding and executing a function. Design is crucial. There is a gap that is not covered by Kaggle experience or starting with a ready-made data set. The instructor focuses on covering general strategies that you can take away as questions you can ask about your upcoming project, such as how to identify Machine learning projects, how to structure a machine learning project for success.
1) Review of Neural Network Theory (Greg)
2) Intro to Apache Spark Machine Learning (Sujee)
3) Lab 1: Set up the environment, Run first set of SparkML commands (Sujee)
4) Intro to Neural Networks with BigDL (Sujee)
5) Lab 2: Set up BigDL environment, apply NNs to Machine Learning tasks.
~ Lunch ~
6) Intro to Convolutional Neural Networks (Alex)
7) Lab 3: Image Recognition with BigDL / LeNet and MNIST dataset
8) Intro to Transfer Learning for Image Recognition (Alex)
9) Lab 4: Transfer Learning using BigDL/Spark and Inception Model.
-Coffee Break –
10) Design of Convolutional Neural Networks (CNNs) and their History (Greg)
11) Lab 5: Modeling Design with SparkML.
Go to TensorFlow Playground (http://playground.tensorflow.org/) to try setting some neural net parameters and training them on different data set.
BEFORE THE CLASS, PREPARATIONS:
• For fun, play around with some neural nets at the TensorFlow Playground (http://playground.tensorflow.org). This will be covered in the class as well.
• You are invited to submit a description of your upcoming machine learning projects or vertical. The instructor will review and may try to incorporate some ideas in the class. Through the meetup site, on the left margin, use the [contact] button.
8:00 – 8:30 arrive, register, coffee, network
8:30 – 10:00 lecture / lab
15 min break, coffee
10:15 – 11:30 lecture / lab
45 min break for lunch
12:15 -1:45 lecture / lab
15 min break, coffee, small snacks
2:00 – 3:30 lecture / lab
15 min break, coffee, small snacks
3:45 – 6:00 lecture / lab
15 min Q&A
ABOUT the Instructors:
Sujee Maniyam (https://www.linkedin.com/in/sujeemaniyam) is a seasoned Big Data practitioner and founder of Elephant Scale. He teaches and consults in Big Data technologies (Hadoop, Spark, NoSQL and Cloud) and Data Science. He is an open source contributor and author of ‘Hadoop illuminated’ (an open-source book on Hadoop) and ‘HBase Design Patterns’. Sujee is a frequent speaker at various conferences and meetups. He also advises and mentors various firms.
Contact : firstname.lastname@example.org
Github : https://github.com/sujee
Alex Kalinin (https://www.linkedin.com/in/alexkalinin) leads AI/Machine Learning team at Sizmek – the largest independent buy-side advertising platform. The team develops cutting edge conversion models, recommender systems, models for automatic A/B testing, and many others. These models are applied in real time at scale, with billions of requests processed per day. Previously, Alex worked in both startups and large companies. While at home.ai he led the team to develop home automation algorithms leveraging computer vision and convolutional neural networks. At Yahoo he led the development of the large-scale user acquisitions and analytics system supporting rapid growth of Yahoo Games business. Alex holds MS in Physics degree, and published several papers on image recognition and pattern detection.
Greg Makowski (https://www.linkedin.com/in/gregmakowski/) has been deploying data mining models for 25 years as the “neural net guy” at American Express/Epsilon. He has developed the analytic internals and automation for 6+ enterprise software systems or SaaS systems. His first convolutional neural net was trained in 1991, a Time Delay Neural Net for speech recognition. Vertical experience includes financial services (credit card, retail banking, bond pricing, ACH payments, fraud detection, customer relationship management (mail, phone, email, banner), retail supply chain among others. He always has something to learn from everybody.