Aws Machine Learning A Complete Guide With Python

Hands on AWS ML SageMaker Course with Practice Test. Join Live Study Group Q&A!

Last updated 2022-01-10 | 4.6

- You will gain first-hand experience on how to train
- optimize
- deploy
- and integrate ML in AWS cloud
- AWS Built-in algorithms
- Bring Your Own
- Ready-to-use AI capabilities
- Complete Guide to AWS Certified Machine Learning – Specialty (MLS-C01)

What you'll learn

You will gain first-hand experience on how to train

optimize

deploy

and integrate ML in AWS cloud

AWS Built-in algorithms

Bring Your Own

Ready-to-use AI capabilities

Complete Guide to AWS Certified Machine Learning – Specialty (MLS-C01)

Includes a high-quality Timed practice test (a lot of courses charge a separate fee for practice test)

* Requirements

* Familiarity with Python
* AWS Account - I will walk through steps to setup one
* Basic knowledge of Pandas
* Numpy
* Matplotlib
* Be an active learner and use course discussion forum if you need help - Please don't put help needed items in course review

Description

Learn about cloud based machine learning algorithms, how to integrate with your applications and Certification Prep

*** NEW Labs - A/B Testing, Multi-model endpoints ***

*** NEW section Emerging AI Trends and Social Issues. How to detect a biased solution, ensure model fairness and prove the fairness ***

*** New Endpoint focused section on how to make SageMaker Endpoint Changes with Zero Downtime ***

*** Lab notebook now use spot-training as the default option. Save over 60% in training costs ***

*** NEW: Nuts and Bolts of Optimization, quizzes ***

*** All code examples and Labs were updated to use version 2.x of the SageMaker Python SDK ***

*** Anomaly Detection with Random Cut Forest - Learn the intuition behind anomaly detection using Random Cut Forest. With labs. ***

*** Bring Your Own Algorithm - We take a behind the scene look at the SageMaker Training and Hosting Infrastructure for your own algorithms. With Labs ***

*** Timed Practice Test and additional lectures for Exam Preparation added

Welcome to AWS Machine Learning Specialty Course!

I am Chandra Lingam, and I am your instructor

In this course, you will gain first-hand SageMaker experience with many hands-on labs that demonstrates specific concepts

We start with how to set up your SageMaker environment

If you are new to ML, you will learn how to handle mixed data types, missing data, and how to verify the quality of the model

These topics are very important for an ML practitioner as well as for the certification exam

SageMaker uses containers to wrap your favorite algorithms and frameworks such as Pytorch, and TensorFlow

The advantage of a container-based approach is it provides a standard interface to build and deploy your models

It is also straightforward to convert your model into a production application

In a series of concise labs, you will in fact train, deploy, and invoke your first SageMaker model

Like any other software project, ML Solution also requires continuous improvement

We look at how to safely incorporate new changes in a production system, perform A/B testing, and even rollback changes when necessary

All with zero downtime to your application

We then look at emerging social trends on the fairness of Machine learning and AI systems.

What will you do if your users accuse your model as racially biased or gender-biased? How will you handle it?

In this section, we look at the concept of fairness, how to explain a decision made by the model, different types of bias, and how to measure them

We then look at Cloud security – how to protect your data and model from unauthorized use

You will also learn about recommender systems to incorporate features such as movie and product recommendation

The algorithms that you learn in the course are state of the art, and tuning them for your dataset is especially challenging

So, we look at how to tune your model with automated tools

You will gain experience in time series forecasting

Anomaly detection and building custom deep learning models

With the knowledge, you gain here and the included high-quality practice exam, you will easily achieve the certification!

And something unique that I offer my students is a weekly study group meeting to discuss and clarify any questions

I am looking forward to seeing you!

Thank you!

Who this course is for:

This course is designed for anyone who is interested in AWS cloud based machine learning and data science
AWS Certified Machine Learning - Specialty Preparation

Course content

26 sections • 245 lectures

Introduction and Housekeeping 14 lectures

Downloadable Resources Preview 00:07

The following downloadable resources are available as part of this lecture:

1. AWS SageMaker Course Introduction.pdf

2. AWS Certified Machine Learning Specialty-Preparation.pdf

3. Gap-Analysis.xlsx

4. AWS Housekeeping.pdf

5. 2020 Benefits of Cloud Computing.pdf

Introduction Preview 02:49

Introduction to AWS Machine Learning Course, Topics Covered, Course Structure

Increase the speed of learning Preview 00:37

Overview - AWS Machine Learning Specialty Exam Preview 09:05

Exam - Gap Analysis Preview 00:24

Preparation - AWS Machine Learning Specialty Exam Preview 04:21

Lab - AWS Account Setup, Free Tier Offers, Billing, Support Preview 07:00

How to set up an AWS account

Different free tier offers from AWS

How to view the charges accrued in your account, and

How to contact AWS support if you need help

Lab - Billing Alerts, Delegate Access Preview 08:10

How to delegate billing access to other authorized users in our account

Configure free tier usage alerts

Set up billing alerts using Cloud Watch and AWS Budget

Lab - Configure IAM Users, Setup Command Line Interface (CLI) Preview 11:30

Configure IAM users required for this course

Set up the AWS command-line tool in your laptop and set the access key credentials.

[Optional] Total Cost of Ownership between On-Premises and Cloud Preview 00:15

Benefits of Cloud Computing Preview 08:39

AWS Global Infrastructure Overview Preview 10:31

Security is Job Zero | AWS Public Sector Summit 2016 by Steve Schmidt Preview 00:04

Weekly AWS Study Group Session Preview 00:36

SageMaker Housekeeping 7 lectures

Downloadable Resources Preview 00:05

Following Downloadable Resources are available in this lecture:

1. Source Code Setup Document

2. Introduction to Machine Learning and Concepts Document

3. usa_airpassengers_numeric.xlsx

Lab - S3 Bucket Setup Preview 02:52

Lab - Setup SageMaker Notebook Instance Preview 02:49

Lab - Source Code Setup Preview 02:25

Kaggle Data Setup Preview 00:18

SageMaker Console looks different from the course videos - Why? Preview 00:38

How to download Kaggle data with code? Preview 00:13

Machine Learning Concepts 5 lectures

Introduction to Machine Learning, Concepts, Terminologies Preview 10:23

Data Types - How to handle mixed data types Preview 12:41

Lab - Python Notebook Environment Preview 10:33

Lab - Working with Missing Data Preview 09:35

Lab - Data Visualization - Linear, Log, Quadratic and More Preview 04:38

Model Performance Evaluation 14 lectures

Model Performance Preview 00:11

Downloadable Resources Preview 00:10

Following Downloadable Resources are available in this lecture:

Model Performance Evaluation Presentation

For exercises in this section, get the latest code from GitHub

https://github.com/ChandraLingam/AmazonSageMakerCourse

If you need help, please refer to SageMaker House Keeping section on how to get the latest code

Introduction Preview 03:26

Lab - Regression Model Performance Preview 09:58

Lab - Binary Classifier Performance Preview 08:00

Lab - Binary Classifier - Confusion Matrix Preview 06:55

Lab - Binary Classifier - SKLearn Confusion Matrix Preview 03:18

Binary Classifier - Metrics Definition Preview 03:52

Binary Classifier - Metrics Calculation Preview 04:26

Question - Why not Model 1? Preview 00:41

Binary Classifier - Area Under Curve Metrics Preview 09:39

Lab - Multiclass Classifier Preview 12:35

Model Performance Preview 00:09

Model Performance Evaluation

SageMaker Service Overview 9 lectures

Downloadable Resources Preview 00:10

How is AWS SageMaker different from other ML frameworks? Preview 01:12

Introduction to SageMaker Preview 04:54

Instance Type and Pricing Preview 10:20

Save Money on SageMaker Usage Preview 02:28

DataFormat Preview 11:12

SageMaker Built-in Algorithms Preview 09:35

Popular Frameworks and Bring Your Own Algorithm Preview 05:24

Infrastructure, Pricing, Support - Review Preview 00:10

SageMaker Service and SDK Changes 6 lectures

Downloadable Resources Preview 00:08

Model Training using Console Preview 08:02

Model Training using Python SDK Preview 08:19

Incremental Training Preview 00:48

Lab - Review the SageMaker console for Training Job Preview 00:12

SageMaker Training

XGBoost - Gradient Boosted Trees 25 lectures

Downloadable Resources Preview 00:02

Introduction to XGBoost Preview 08:52

"The XGBoost (eXtreme Gradient Boosting) is a popular and efficient open-source implementation of the gradient boosted trees algorithm".

https://docs.aws.amazon.com/sagemaker/latest/dg/xgboost.html

Introduction to XGBoost and how it compares to the Linear Model, Decision Tree, and Ensemble Methods

Lab - Data Preparation Simple Regression Preview 05:06

Let's compare performance of XGBoost and Linear Model using a simple regression dataset

Lab - Training Simple Regression Preview 12:25

Train model using XGBoost and Linear Regression. Evaluate performance using Plots, Residual Histograms and RMSE metrics

Lab - Data Preparation Non-linear Data set Preview 02:39

Let's compare performance of XGBoost and Linear Model using a non-linear data set

Lab - Training Non-linear Data set Preview 04:47

Exercise - Improving quality of predictions Preview 00:12

Lab - Data Preparation Bike Rental Regression Preview 08:24

In this lab, let's look at Bike Rental demand forecasting problem. This is an old competition problem from Kaggle: https://www.kaggle.com/c/bike-sharing-demand/data. To download data files, you need to register with Kaggle (it's free).

Lab - Train Bike Rental Regression Model Preview 06:09

In this lab, let's train our model for forecasting hourly bike rental counts. This is a complex non-linear data set that has seasonality, trend and several factors that impact rentals. Evaluate quality of predictions using Plots, Residual Histograms, RMSE and RMSLE metrics. Finally, submit the results at Kaggle for test data.

Lab - Train using Log of Count Preview 04:14

In this lab, let's transform the target using log operation. Log of target can help when the target is a count/integer, it has seasonality and trend. After model predicts the value, we need to apply inverse transform (exp) to get the count back.

ResourceLimitExceeded Error - How to Increase Resource Limit Preview 00:38

Lab - How to train using SageMaker's built-in XGBoost Algorithm Preview 07:36

In this lab, let's train bike rental model on SageMaker's built-in XGBoost Algorithm. We will walk through the fours steps for using a SageMaker algorithm

Q&A: How does SageMaker built-in know the target variable? Preview 00:28

Lab - How to run predictions against an existing SageMaker Endpoint Preview 04:29

In this lab, let's look at the steps involved in connecting to an existing SageMaker endpoint, security of an endpoint, how to send multiple observations in each call.

SageMaker Endpoint Features Preview 05:41

Let's look at key benefits of a managed Endpoint. SageMaker takes care of automatic replacement of unhealthy instances, AutoScaling infrastructure based on workload, hosting multiple versions of model behind an endpoint, and metrics published to CloudWatch.

SageMaker Spot Instances - Save up to 90% for training jobs Preview 01:34

Lab - Multi-class Classification Preview 05:41

In this lab, let's look at multi-class classification using XGBoost.

Lab - Binary Classification Preview 06:21

In this lab, let's look at a how to perform Binary Classification using XGBoost. We will use the diabetes data set in this lab

Exercise - Improve Data Quality in Diabetes dataset Preview 00:19

Question on Diabetes Data Quality Improvement Preview 00:15

Question on Diabetes model - is group mean on target the right approach? Preview 00:04

HyperParameter Tuning, Bias-Variance, Regularization (L1, L2) Preview 11:07

In this lecture, let's look at important XBoost Hyperparameters. We will also look at Bias, Variance, Regularization (L1, L2), and Automatic Tuning

Exercise - Mushroom Classification Preview 00:15

Quiz - XGBoost

Underfitting, Overfitting

Invoke Model Endpoint From External Clients 10 lectures

Install SageMaker SDK, GIT Client, Source Code, Security Permissions Preview 00:13

IAM users for the lab Preview 00:09

Integration Overview Preview 02:32

Lab - Client to Endpoint using SageMaker SDK Preview 09:26

Lab - Client to Endpoint using Boto3 SDK Preview 03:50

Microservice - Lambda to Endpoint - Payload Preview 03:24

Lambda UI Changes Preview 00:14

Lab - Microservice - Lambda to Endpoint Preview 09:09

API Gateway - UI Changes Preview 00:06

Lab - API Gateway, Lambda, Endpoint Preview 10:34

Endpoint Changes with Zero Downtime 7 lectures

Downloadable Resources Preview 00:01

[Repeat] Endpoint Features, Monitoring and AutoScaling Preview 05:41

How to handle changes to production system? Preview 07:54

Lab - A/B Testing Multiple Production Variants Preview 11:53

Lab – Multi-model Endpoint Preview 08:15

Run Models at the Edge Preview 01:22

Endpoints

Emerging AI Trends and Social Issues 5 lectures

Downloadable Resources Preview 00:01

Is AI Biased? Preview 07:40

Tools to Detect Bias - Clarify, Experiments, Model Monitor, Augmented AI Preview 07:41

And Some More Tools Preview 00:16

Emerging AI Trends and Social Issues

Cloud Security and Access Management 12 lectures

Introduction Preview 00:14

Shared Responsibility Model, Compliance, Delegation, Federation Preview 08:33

Credentials, MFA, Identity-based, Resources-based Policy Preview 07:24

Inline and Managed Policy, Amazon Resource Naming (ARN) Convention Preview 08:30

Principal, Effect, Action, Resource, Not Clause Preview 06:55

Conditional Access, Implicit Deny, Explicit Allow and Deny, Permission Boundary Preview 07:42

IAM Roles, Cross-account access options Preview 06:17

Federation, SSO, SAML, Active Directory, AWS Organizations, Cognito Preview 06:16

Lab - Identity-based policy, Implicit Deny, Explicit Allow Preview 05:01

Lab - Policy Generator, Managed Policy, Versions, Groups Preview 05:40

Lab - Resource-based policy, Policy Generator, Principals Preview 05:14

Cloud Security

Principal Component Analysis (PCA) 13 lectures

Normalization and Standardization Preview 00:23

Downloadable Resources Preview 00:07

Introduction to Principal Component Analysis (PCA) Preview 05:49

"PCA is an unsupervised machine learning algorithm that attempts to reduce the dimensionality (number of features) within a dataset while still retaining as much information as possible. This is done by finding a new set of features called components, which are composites of the original features that are uncorrelated with one another. They are also constrained so that the first component accounts for the largest possible variability in the data, the second component the second most variability, and so on."

https://docs.aws.amazon.com/sagemaker/latest/dg/pca.html

PCA Demo Overview Preview 01:16

Demo - PCA with Random Dataset Preview 03:29

Demo - PCA with Correlated Dataset Preview 05:26

Cleanup Resources on SageMaker Preview 00:28

Demo - PCA with Kaggle Bike Sharing - Overview and Normalization Preview 03:51

Demo - PCA Local Mode with Kaggle Bike Train Preview 03:30

Demo - PCA training with SageMaker Preview 04:22

Demo - PCA Projection with SageMaker Preview 02:42

Exercise : Kaggle Bike Train and PCA Preview 00:23

Summary Preview 01:22

Recommender Systems - Factorization Machines 7 lectures

Recommender System Preview 01:10

Downloadable Resources Preview 00:07

Introduction to Factorization Machines Preview 05:59

"A factorization machine is a general-purpose supervised learning algorithm that you can use for both classification and regression tasks. It is an extension of a linear model that is designed to capture interactions between features within high dimensional sparse datasets economically. For example, in a click prediction system, the factorization machine model can capture click rate patterns observed when ads from a certain ad-category are placed on pages from a certain page-category. Factorization machines are a good choice for tasks dealing with high dimensional sparse datasets, such as click prediction and item recommendation."

https://docs.aws.amazon.com/sagemaker/latest/dg/fact-machines.html

MovieLens Dataset Preview 00:08

Demo - Movie Recommender Data Preparation Preview 10:35

Demo - Movie Recommender Model Training Preview 05:34

Demo - Movie Predictions By User Preview 07:10

Model Optimization and HyperParameter Tuning 8 lectures

Downloadable Resources Preview 00:02

Introduction to Hyperparameter Tuning Preview 06:11

"Amazon SageMaker automatic model tuning, also known as hyperparameter tuning, finds the best version of a model by running many training jobs on your dataset using the algorithm and ranges of hyperparameters that you specify. It then chooses the hyperparameter values that result in a model that performs the best, as measured by a metric that you choose."

https://docs.aws.amazon.com/sagemaker/latest/dg/automatic-model-tuning.html

Lab: Tuning Movie Rating Factorization Machine Recommender System Preview 18:05

Lab: Step 2 Tuning Movie Rating Recommender System Preview 05:00

HyperParameter, Bias-Variance, Regularization (L1, L2) [Repeat from XGBoost] Preview 11:07

Nuts and Bolts of Optimization Preview 00:31

Model Optimization

Model Optimization - related question Preview 00:02

Time Series Forecasting - DeepAR 12 lectures

Downloadable Resources Preview 00:06

Introduction to DeepAR Time Series Forecasting Preview 09:47

"The Amazon SageMaker DeepAR forecasting algorithm is a supervised learning algorithm for forecasting scalar (one-dimensional) time series using recurrent neural networks (RNN)"

https://docs.aws.amazon.com/sagemaker/latest/dg/deepar.html

DeepAR Training and Inference Formats Preview 09:48

Working with Time Series Data, Handling Missing Values Preview 09:58

Demo - Bike Rental as Time Series Forecasting Problem Preview 11:43

Demo - Bike Rental Model Training Preview 07:21

Demo - Bike Rental Prediction Preview 04:50

Demo - DeepAR Categories Preview 06:10

Demo - DeepAR Dynamic Features Data Preparation Preview 06:34

Demo - DeepAR Dynamic Features Training and Prediction Preview 03:05

Summary Preview 01:15

Question: How to train a model for different products using DeepAR? Preview 00:53

Anomaly Detection - Random Cut Forest 4 lectures

Downloadable Resources Preview 00:02

Introduction to Random Cut Forest and Intuition Behind Anomaly Detection Preview 10:10

Lab - Taxi Passenger Traffic Analysis (AWS Provided Example) Preview 08:53

Lab - Auto Sales Analysis Preview 05:54

Artificial Intelligence (AI) Services 16 lectures

Downloadable Resources Preview 00:17

Amazon AI Free Tier

Amazon AI Services are now part of the 12-month free tier.

However, there are some exclusions. For example, Comprehend service custom model training is not part of free-tier.

The good news is you can gain hands-on experience on most of these capabilities for free

Please check the new housekeeping video: 2020 AWS Account Setup, Free Tier Offers, Billing, Support

The following downloadable resources are available in this lecture:

AWS AI Services.pdf

Lab Instructions Preview 00:13

1. For the labs in the AI Section, login to AWS Management Console using my_admin (administrative) account that was setup in House Keeping Lectures.

2. Download Latest Source Code from Git:

From your SageMaker Notebook's AmazonSageMakerCourse folder, Run the command:

git pull

If you need help, please review SageMaker Housekeeping section of this course.

1. Introduction Preview 03:15

2.1 Amazon Transcribe and Lab Preview 05:32

2.2 Amazon Transcribe and Lab Preview 06:34

3. Amazon Translate Preview 04:29

Translate - Practical Scenario Preview 00:12

4.1 Amazon Comprehend Preview 05:42

Pricing Comprehend Preview 00:32

4.2 Amazon Comprehend Preview 05:00

4.3 Amazon Comprehend training Preview 08:35

5. Amazon Polly Preview 04:16

6. Amazon Lex Preview 06:48

7. Amazon Rekognition Preview 08:21

8. Amazon Textract & Summary Preview 03:02

AI Services Quiz

S3 Data Lake Architecture - Data Consolidation 20 lectures

Downloadable Resources Preview 00:02

Following Downloadable Resources are available in this lecture:

AWS Data Lake.pdf

Lab Instructions Preview 00:13

Introduction to Data Lake Preview 10:28

In this section, let’s look at how a Data Lake can help you streamline data management

We will focus on the general concept and specific AWS products that can help you build a data lake.

Topics Covered:

1. Data Lake Vs. Data Warehouse

2. Storage Solutions: S3, Glacier

3. Data Ingestion and Migration: Kinesis, Storage Gateway, Snowball, Snowmobile

4. Meta Data Catalog: Glue and Custom

Kinesis - Streaming and Batch Processing Preview 05:23

Data Formats and Tools for Data Format Conversion Preview 08:33

In this lecture, let’s compare popular data formats for storing your data in a data lake and tools that you can use for format conversion.

You will gain insight into Row and Columnar storage formats and popular storage formats like CSV, TSV, JSON, Parquet, ORC, and Avro.

We will review popular approaches for format conversion using EMR Hive, Spark, Glue ETL

In-Place Analytics and Portfolio of Tools Preview 05:02

One of the exciting capabilities of S3 based data lake architecture is in-place querying.

You can directly run SQL queries on your data stored in S3 without having to load in a database or data warehouse.

We will review Athena and Redshift Spectrum for in-place querying.

For streaming data, we will Kinesis Data Analytics

Your data lake needs to be flexible to support a wide variety of current and future tools.

Here is the thing, S3 is one of the most popular products in the cloud, and a lot of tools natively support S3.

We will review EMR Hadoop based tools like Hive, Spark, SageMaker Machine Learning, AWS Artificial Intelligence Services, QuickSight Business Intelligence Service, Redshift Data Warehouse and Lambda

Monitoring and Optimization Preview 06:27

In this lecture, let’s review tools for monitoring and optimization of your data lake.

Monitoring and Auditing services reviewed: CloudWatch, CloudWatch Logs, CloudTrail

For Optimization, we will look at:

S3 Storage Classes: Standard, Standard-IA Infrequent-Access, Intelligent-Tiering, Glacier, Glacier Deep Archive

S3 Lifecycle Management

S3 Storage Classes Analysis

Security and Protection Preview 06:36

S3 Access Control using Resource and User-based policies

Data Encryption using S3 Server Side Encryption, Client-Side Encryption

Encryption Key Management using Key Management Service (KMS)

Default Master Key and Customer Master Key (CMK)

S3 durability to protect against Corruption and Data Loss

S3 Versioning to protect against malicious and accidental deletes

Multi-Factor Authentication (MFA)

Cross-Region Replication (CRR)

Object Tagging and Tag-based Security policies

Quiz - Data Lake

Lab Instructions - Glue Data Catalog Preview 02:49

Lab – Glue Data Catalog Preview 08:31

In-place Querying of files stored in S3

•Store file in S3

•Collect metadata with Glue Crawler

•Run Query using Athena

Lab Instructions – Athena In-place Querying Preview 01:25

Lab - Query with Athena Preview 02:01

In-place Querying of files stored in S3

•Run Query using Athena

Lab - Glue ETL - Convert format to Parquet Preview 04:43

Use Glue ETL to convert files to Parquet format

•Glue automates process of ETL script generation, scheduling and execution

•Glue ETL provisions required Apache Spark infrastructure to run the job

Lab - Query Amazon Customer Reviews with Athena Preview 05:06

Query Amazon Customer Reviews Public Dataset using Athena

•Create table definition (instead of using Glue Crawler)

•Update catalog with partition

•Query using Athena

Lab – Sentiment of the Customer Review Preview 06:06

Find Sentiment of the customer review using Comprehend AI Service

With Athena, Query the reviews using sentiment

Lab - Query Sentiment of Customer Reviews using Athena Preview 04:16

Lambda UI Changes Preview 00:11

Lab – Serverless Customer Review Solution Part 1 Preview 09:45

In this lab, we are going to build a solution that accepts customer reviews at unlimited scale, process the sentiment and deliver results to S3

The entire solution is serverless – except for a small client that will generate reviews.

This solution will use:

· Python Client as a streaming data source

· Kinesis Firehose for ingesting streaming data,

· Lambda function for transformation to add a sentiment,

· S3 Data Lake as a destination for the transformed data.

· Glue and Athena for a near real-time data querying

Lab – Serverless Customer Review Solution Part 2 Preview 07:51

Deep Learning and Neural Networks 21 lectures

ReadMe and Downloadable Resources Preview 00:29

Lab Instructions Preview 00:13

Concepts - Gradient Descent, Loss Function for Regression Preview 14:11

In this section, we will learn about Gradient Descent Optimizer, Loss Functions, Learning Rate, Variants of Gradient Descent Optimizers like RMSProp, Adagrad, Adam.

You will also learn the differences between batch, stochastic, and mini-batch modes of gradient calculation.

We will use a regression model for the discussion

Concepts - Gradient Descent, Loss Function for Classification Preview 10:02

In this section, we will learn about Gradient Descent Optimizer, Loss Functions, Learning Rate, Variants of Gradient Descent Optimizers like RMSProp, Adagrad, Adam.

You will also learn the differences between batch, stochastic, and mini-batch modes of gradient calculation.

We will use a classification model for the discussion

Neural Networks and Deep Learning Preview 07:35

In this lecture, we will build on the concepts to build a neural network: neurons, activation functions, input layer, hidden layers, an output layer, and optimization strategies.

We will learn about why deep learning is popular in some of the domains

We will look at different ways in which you can structure your network for regression, binary classification, and multi-class classification problems.

Finally, we will review popular network architectures: fully connection general-purpose neural network, Convolutional Neural Networks and Recurrent Neural Networks