Scribd
Upload
29 views

Machine Learning Basics Infographic With Algorithm Examples

This document provides an overview of Azure Machine Learning and how it can be used to perform various types of data analysis tasks through different machine learning algorithms. It describes how Azure ML works by training models on current data to find patterns that can then be applied to make predictions on future data. As an example, it walks through the steps of using regression analysis on car rental data to accurately predict future product demand. It also includes a list of algorithm modules that can be used for tasks like regression, anomaly detection, clustering, and classification.

Uploaded by

Yacoov Tovim
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
29 views

Machine Learning Basics Infographic With Algorithm Examples

This document provides an overview of Azure Machine Learning and how it can be used to perform various types of data analysis tasks through different machine learning algorithms. It describes how Azure ML works by training models on current data to find patterns that can then be applied to make predictions on future data. As an example, it walks through the steps of using regression analysis on car rental data to accurately predict future product demand. It also includes a list of algorithm modules that can be used for tasks like regression, anomaly detection, clustering, and classification.

Uploaded by

Yacoov Tovim
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

DIG DEEP WITH

AZURE
MACHINE
LEARNING

Use data analysis to take your business to a whole new level.

Microsoft Azure Machine Learning simplifies data analysis and


empowers you to find the answers your business needs.

The question isn’t whether you can find the answers.


The question is how.

So, what do you want to find out?


I WANT TO:

Regression Estimate product demand

Forecast the future by Predict


Predict sales figures
estimating the relationship Values
between variables. Analyze marketing returns

Predict credit risk


Find Anomaly Detection
Detect fraud
Unusual Identify and predict rare or
Occurrences unusual data points.
Catch abnormal equipment readings

Perform customer segmentation


Clustering Discover
Separate similar data points Structure Predict customer tastes
into intuitive groups.
Determine market price

Predict Classification
Categories Identify what category new
information belongs in.

Predict Between Predict Between


Two Categories Several Categories
Two-Class Classification Multi-Class Classification

Answers simple Answers complex


two-choice questions, like questions with multiple
yes-or-no, true-or-false. possible answers.

Is this tweet What is the mood


positive? of this tweet?

Will this customer Which service will this


renew their service? customer choose?

Which of two coupons Which of several promotions


A/B/C
draws more customers? draws more customers?

Azure Machine Learning works by teaching the


software to find patterns in the current data so that it
can seek out the patterns in future data.

Let’s say you rent cars.


How can you accurately predict
demand for your product?
FOR THAT YOU NEED REGRESSION ANALYSIS

STEP 01 STEP 05
Get the data. Predict future demand.
Car rental could spike Use the model to forecast
depending on time of day, future spikes and
holidays, weather, etc. shortfalls in demand.

STEP 02 STEP 04
Prepare the data. Score and evaluate the model.
Clean data, combine Test the model’s ability to
datasets, and prepare predict the original data, and
it for analysis. evaluate its success.

STEP 03
Train the model.
Feed the information
into the machine to
teach it what to expect.

Find out how to do this and more with #AzureML.


Visit us at https://studio.azureml.net/

ALGORITHM MODULE OPTIONS

Regression

Ordinal Regression Poisson Regression Fast forest quantile Linear Regression


Data in rank ordered Predicts event regression Fast training, linear
categories counts Predicts a distribution model
Example > Example > Example > Example >

Bayesian Linear Neural Network Decision Forest Boosted Decision


Regression Regression Regression Tree Regression
Linear model, small Accurate, long Accurate, fast Accurate, fast training times,
data sets training times training times large memory footprint
Example > Example > Example > Example >

Anomaly Detection

One Class SVM PCA-Based Anomaly Detection


Under 100 features, aggressive boundary Fast training times
Example > Example >

Clustering

K-Means
Unsupervised learning
Example >

Two-Class Classification

Two-class SVM Two-class averaged Two-class Bayes


Under 100 features, perceptron point machine
linear model Fast training, linear Fast training, linear
model model
Example > Example > Example >

Two-class Two-class logistic Two-class boosted


decision forest regression decision tree
Accurate, fast Fast training, linear Accurate, fast training,
training model large memory footprint
Example > Example > Example >

Two-class Two-class Two-class


decision jungle locally deep SVM neural network
Accurate, small Under 100 Accurate, long
memory footprint features training times
Example > Example > Example >

Multiclass Classification

Multiclass logistic Multiclass neural Multiclass


regression network decision forest
Fast training times, Accuracy, long Accuracy, fast
linear model training times training times
Example > Example > Example >

Multiclass One-v-all
decision jungle multiclass
Accuracy, small Depends on the
memory footprint two-class classifier
Example > Example >

You might also like