The Machine Learning exam is rather difficult, as discussed previously. The starting point would be the acloud.guru Machine Learning course or Linux Academy courses. Additionally is the training offered by AWS. A chunk of Machine Learning is data and data preparation, so please see my links from Big Data.

Here is a collection of links I put together which helped me with studying for the exam.

General Topics

Handling imbalances in Data
Learning Rate
Neural Networks
Common Machine Learning Algorithms
Another Resource on Machine Learning Algorithms

Machine Learning Concepts

Formulating the Problem
Regression
Regression Model Insights
The Machine Learning Process
Machine Learning Key Concepts
Cross Validation
Splitting Training Data
Training Parameters
Training Faster with Sagemaker Linear Learner
Multiclass Model Insights
Managing Machine Learning Projects Whitepaper
SageMaker Blog
Underfitting and Overfitting
Machine Learning Models
Binary Model Insights
One Hot Encoding

Data

Glue
Glue Crawler
Athena
SparkML
KPL
Kinesis Data Firehose
Kinesis PutRecord

AWS Machine Learning - SageMaker

Data Formats
SageMaker Batch
SageMaker Docker Registry
Built-in Algorithms
Elastic Inference
Elastic Inference
Inference Pipeline Containers
Validating a Model
Training Metrics
CloudTrail
AutoScaling
SageMaker with Step Functions
Hosting Model
SageMaker and IAM
Polly

SageMaker Machine Learning Implementations

Semantic Segmentation
Seq-to-Seq
K-Means Linear Learner
Linear Learner Tuning
BlazingText BlazingText InputOutput
LDA
Factorization Machines
Random Cut Forest
K Nearest Neighbor
Image Classification
Object2Vec
Object Detection
PCA
DeepAR
XGBoost
XGBoost Tuning
XGBoost Parameters
Neural Topic Model

SageMaker TensorFlow Framework

TensorFlow

SageMaker Hyperparameter Tuning

Creating Hyperparameters Tuning Job
Automated Tuning
Hyperparameter Tuning Job
Hypertunning
Image Classification Hyperparameters

Business Intelligence

QuickSight
Chart types

I passed the AWS Certified Machine Learning Speciality Exam on Monday. That makes my 10th AWS certification in the last 18 months.

The Machine Learning Specialty certification is unlike any of the other exams from AWS. The exam doesn’t just focus on AWS specifics but covers a wide range of Machine Learning topics. The exam blueprint provides a basis of this coverage.

The exam is probably the hardest of the 10 I’ve taken to date. The entire exam, I thought I know the material, but I don’t think I know it well enough to pass the exam. My score was good, and it satisfying to add this certification. For the Machine Learning exam, I put in well over 200 hours over the last six months and over 80 hours the four weeks before sitting the exam. Definitely think the Big Data Certification helped on the data preparation sections.

They’re a bunch of links I will share later this week, which I studied. In addition to all the reading, I did acloud.guru’s AWS Certified Machine Learning - Speciality, which provides 40% of the material required to pass the exam. The rest of the exam requires detailed knowledge of Machine Learning. I followed the learning track recommended by AWS for Data Scientist. I also did several sections from Linux Academy Machine Learning, including the great section explaining PCA. Lastly, I took the AWS practice exam. I did look at Whizlabs but was somewhat disappointed in their practice tests.

In 2020, I hope to get a project which will allow me to leverage Machine Learning in SageMaker to solve a complex customer problem.