Course Launch - Scaling and Accelerating Machine Learning Models

Welcome to "Scaling and Accelerating Machine Learning Models" course. This is going to be complete hands on course covering entire Machine Learning Life cycle starting from Data Collection, Data Engineering till model deployment. This will also include topics on cloud based Data and AI services

Who is this course for?

Course is for aspiring as well as experienced data scientist, data engineers, data analyst, data architects and software engineers

For some of the modules knowledge of python might be required but there will be enough topic to consume for specific roles, who do not have python experience as well. Basic knowledge on Machine Learning is must to have for some of the modules

If you are new to data science and want to learn on life cycle of machine learning projects, would recommend watching my End to End Machine Learning Videos

Watch atleast first video to get overall picture of ML lifecycle. Ideal is to spend time to go through most of the videos in End to End Machine Learning to some extent

If you are completely new to machine learning, check appendix section on link from where you can learn data science

Course Curriculum

Scaling ML Pipline

The course is purely focused on Scaling your machine learning pipeline on to distributed ecosystem as well as hardware accelerators like GPU and TPU

Accelerating ML projects to generate faster insights and value

Machine Learning jobs are compute intensive as well as time intensive and faster iterations are key to generate business value. Accelerating machine learning comes from automation and part of the course will cover techniques for automating pieces of machine learning pipeline

Quick Overview of Course and Modules covered

Module 1 - Apache Spark

• Overview of Apache Spark
• Spark for Data Engineering - Coming by Feb End
• Spark for Machine Learning
• Spark for Time Series Analysis - Coming by Feb End
• Spark for Deep Learning - Coming by Feb End

Module 2 - TensorFlow

• Overview of TensorFlow
• End to End TensorFlow Deep Learning model on GPU
• Accelerating Machine Learning models on TPU
• Tensorflow Custom graphs on TPU
• TensorFlow Extended for Data Validation and Analysis - Coming by Feb End
• Tensorflow for Time Series Analysis - Coming in March

Module 3 - XGBoost, RAPIDS and Numba

• Overview of RAPIDS
• RAPIDS for Data Analysis on GPU
• RAPIDS for Machine Learning
• Accelerating XGBoost on GPU
• Overview of Numba and Build Machine learning model from scratch using Numba - Coming in March

Module 4 - Cloud Data and AI Services (Coming in March-April Timeframe)

• Overview of Cloud for Data and AI
• Architecture pattern with cloud native technologies (AWS and Google Cloud)
• Amazon Sagemaker
• Google Cloud AI platform
• BigQuery ML
• Cloud API for NLP and Vision
• Azure ML
• Hybrid Cloud for Data

Module 5 - AutoML

• Why do we need AutoML?
• Hands on Demo of prominent AutoML framework like H20 AutoML, AutoViML, Gluon, Auto SKlearn, TPOT etc

Module 6 - Model Deployment (Coming in June)

• Docker and Kubernetes for Data Science
• Architecture pattern for real time streaming
• Deploying models on real time streaming service and containers to handle high velocity transactions

How can One access the course?

You can subscribe to my channel using link below to get notified as I post new videos -

Link to Channel playlist where you can view and play the course

What is the tentative date for completion of entire course?

Some part of the course are already out there and tentative target for uploading the remaining part of the course is by July' 2020. Remember I get only weekend to make my videos and in case if there is further delay I will keep you posted. You can see videos coming out in frequent interval till July' 2020

What is the cost of course?

Course is completely free to learn for all. But understand the course is not free for me as I am spending my time and energy to get this course out for those who are willing to invest their time. Only thing I am requesting is "do not personal message me" on this course. For any doubt or questions use YouTube video comments so others can contribute to answer it as well

Do we need to learn modules in order mentioned?

Every module is independent and you can pick and chose topics of your interest. In case if there is reference to cross module, will provide link within the video

How can we contact you for any clarification?

YouTube comments on specific video if you need additional clarification

How can we help you?

Share this post with your friends, colleagues or office mates so larger community can benefit from it

Is this enough for anyone to get a Job?

No. This alone might not help you. My intention is to make you enterprise ready and not job ready. In fact any course out there today might not be able to promise you a job. Learn and Practice to standout in interview and this is just one platform to learn and practice

Will you be sharing the code used in the Hands on lab?

Yes, for selected modules. For some which are critical to learn by being hands on, I might not as I want to you to practice it. Many times we learn new tools by building and debugging the issue

Link to my github repo where you can find code for selected modules -

Appendix

New to data science, use this learning path to get started