Complete Azure AI Fundamentals (AI-900) Certification Course Overview

hey this is Andrew Brown and I'm bringing you another certification course and this time it's the Azure AI

fundamentals also known as the AI 900 and if you're looking to pass a certification we have everything that

you need here such as Labs lectures and a free practice exam so you can go ASAT exam get that uh certification put on

your resume and Linkedin to go get that job you've been looking for um if you want to support more free courses like

this one the best way is to purchase the additional uh paid materials where you can get access uh to more practice exams

and other resources if you don't know me um I've taught a bit of everything uh here on the cloud that's been with adabs

Azure gcp devops terraform kubernetes you name it I've taught it but uh you know you know the drill here let's get

into it and learn more about the Azure AI fundamentals hey this is Andrew Brown

from exam Pro and we are at the start of our journey here learning about the AI 900 asking the most important question

which is what is the AI 900 so the Azure AI fundamental certification is for those seeking an ml role such as AI

engineer or data scientist the certification will demonstrate if a person can Define and understand Azure

AI services such as competive services and Azure applied AI Services AI Concepts knowledge mining responsible AI

basics of NL pipelines classical ml models autom ml generative AI workloads which is newly added content and Azure

AI studio so you don't need to know super complicated ml knowledge here but it definitely helps to get you through

there so this certification is generally referred to by its course code the AI 900 and it's the natural path for the

Azure AI engineer or Azure data scientist certification this generally is an easy course to pass and it's great

for those new to cloud or ml related technology looking at our road map you might be asking okay well what are the

paths and what should I learn first so here are a few suggested routes if you already have your a900 that's that's a

great starting point before you take your AI 900 if you don't have your a z900 you can jump right into the AI 900

but I strongly recommend you go get that a z900 because it gives you General foundational knowledge it's just another

thing that you should not have to worry about which is just how to use Azure at a fundamental level do you need the

dp900 to take the AI 900 no but a lot of people seem to like to go this route where they want to have that data

Foundation before they move on to the AI 900 because they know that the broad knowledge is going to be useful they so

it's app pairing that you see a lot of people getting the AI 900 and the dp900 together for the AI 900 the path is a

little bit more clear it's either going to be data scientists or AI engineer so for the AI engineer you have to know how

to use the AI services in and out for data scientists it's more focused on setting up actual pipelines and things

like that within Azure machine learning so you just have to decide which path is for you the data scientist is definitely

harder than the AI engineer so if you aren't ready for the data science some people like taking the AI engineer first

and then doing the data scientists so this is kind of like a warmup again it's not 100% necessary but it's just based

on your personal learning style and a lot of times people like to take the data engineer after the data scientists

just to round out their complete knowledge now if you already have the a900 and the administrator associate you

can safely go to the data scientist if you want to risk it because this one is really hard so if you've passed the

a1004 you know you're going to probably have a lot more confidence learning up about all the concepts at this level

here but of course it's always recommended to go grab these foundational CTS because sometimes

course materials just do not cover the information and so the obvious stuff is going to get left out okay so moving

forward here how long should you study to pass for the AI 900 if you're entirely new to ml Ai and Cloud

providers such as Azure you should anticipate dedicating around 15 hours to grasp the basics this estimate can vary

base on your familiarity with these concepts for complete beginners the time commitment might extend to 20 to 30

hours for the intermediate level so people that have passed the a900 or dp900 you're looking at around 8 to 10

hours if you have one or more years of experience with Azure or another cloud service provider like a WS or gcp you're

looking at about 5 hours or less the average stunny time is about 8 hours this is where you should be committing

50% of the time to the lecture in labs and 50% for the practice exams the recommended study time is 30 minutes to

an hour a day for 14 days this should get you through it but just don't overstudy and just don't spend too

little time what does it take to pass the exam well you got to watch the lectures and memorize key information do

Hands-On labs and follow along with your own Azure account I'd say that you could probably get away with just watching all

the videos in this one without having to do the labs but again it really does reinforce that information if you do

take the time there is some stuff that is in Azure AI Studio or machine learning you might be wary of launching

instances because we do have to run instances and they will cost money unless you delete the instances after

use resulting in very small costs so if you feel that you're not comfortable with that by just watching the videos

you should be okay but when you get into the associate tier you absolutely have to expect to pay something to learn and

take that risk you want to do paid online practice exams that simulate the real exam as I've mentioned before I do

provide a free practice exam and have paid practice exams that accompany this course that are on my platform exam Pro

and that's how you can help support more of these free courses so can you pass this certification without taking a

practice exam well azzure is a little bit harder if this isn't a WS exam I would say yes but for Azure exams like

AI 900 dp900 and sc900 probably not it's kind of risky I think you should do at least one practice exam or go through

the sample one there's a sample one probably laying around on the Azure website let's take a look at the exam

guide breakdown here and then in the following video we'll look at in more detail so it's broken down into the

following domain so the exam has five domains of questions and each domain has its own waiting which determines how

many questions in a domain that will show up so 15 to 20% will be described Ai workloads and considerations 20 to

25% will consist of describe fundamental principles of machine learning on Azure 15 to 20% will consist of described

features of computer vision workloads on Azure 15 to 20% will be described features of natural language processing

workloads on Azure and 15 to 20% will be described features of generative AI workloads on Azure I want you to notice

it's says describe these domains this is good because that tells you it's not going to be super hard if you start

seeing things that say Beyond describe and identify then you know it's going to be a bit harder so where do you take

this exam well you can take it in person at a test center or online from the convenience of your own home so there's

two popular test centers there's CER aort and there's Pearson view you can also take it at a local test center if

there are nearby locations the term Proctor means a supervisor or person that is monitoring you while you're

taking the exam if I had the the option between in person or online I would always choose the in person because it's

a controlled environment and it's way less stressful online there are many things that can go wrong but it's up to

your personal preference and your situation the passing grade here is 700 out of a th000 so that's around 70% I

would say around because you could possibly fail with 70% because these things work on scaled scoring for

response types there's about 37 to 47 questions and you can afford to get about 10 to 13 questions wrong so some

questions are worth more than one point some questions cannot be skipped and the format of questions can be multiple

choice multiple answer drag and drop and hot area there shouldn't be any case studies for foundational level exams and

there's no penalty for wrong questions so for the duration you get 1 hour that means about 1 minute per question the

time for this exam is 60 Minutes your C time is 90 minutes C time refers to the amount of time that you should take to

allocate for that exam so this includes time to review the instructions read and accept the NDA complete the exam and

provide feedback at the end this certification is going to be valid forever and it does not expire Microsoft

fundamental certifications such as the a900 or ms9900 do not expire as long as the technology is still available or

relevant so we'll proceed to the full exam guide [Music]

now hey this is Andrew Brown from exam Pro and what we've pulled up here is the official exam outline on the Microsoft

website if you want to find this year s you just have to type in AI 900 Azure or Microsoft you should be able to easily

find it the page looks like this what I want you to do is scroll on down because we're looking for the AI 900 set of

guide and from there we're going to scroll on down to the skills measured section and you might want to bump up

the text Azure loves updating their courses with minor updates that don't generally affect the outcome of the

study here but it does get a lot of people worried because they always say well is your course out of date so no

they're just making minor changes because they'll do this like five times a year and so if there was a major

revision what would happen is they would change it so instead of being the AI 900 it would be like the AI 9001 or 9002

similar to how the AI 102 was previously AI 100 but now it's the AI 102 so just watch out for those and if it's a major

revision then yes it would probably need a completely new course so there aren't any major changes with the new update

other than the update for the generative AI workloads on Azure section A couple of name changes and a few things being

removed everything else remains is relatively the same with very minor changes so the concepts and such are

still up to date overall I think the exam is easier than the four so let's go through some of the topics and work our

way through here so describe Ai workloads and considerations so here we're just kind of describing the

generalities of AI so content moderation workloads involve filtering out inappropriate or harmful content from

user generated inputs ensuring a safe and positive user experience personalization workloads analyze user

behavior and preferences to tailor content recommendations or experiences to individual users computer vision

workloads involve the analysis of images and videos to recognize patterns objects faces and actions identify natural

language processing knowledge mining document intelligence and features of generative AI workloads note that these

are all just Concepts you don't need to know how to use the services at a high level then you have the responsible AI

section so Microsoft has these six principles that they really want you to know and they push it throughout all

their AI services so those are the six you'll need to know and they're not that hard to

learn moving on we have described fundamental principles of machine learning on

Azure so here it's just describing regression classification clustering and features of deep learning we have a lot

of practical experience with these in the course so you will understand at the end what these are used for next we have

core machine learning Concepts we can identify features and labels in a data set so that's the data labeling service

describe how training validation data sets are used in machine learning so we'll touch on that describe

capabilities of Automated machine learning automl simplifies building and picking the best models while data and

compute Services provide the power you need for training with Azure machine learning it helps with managing and

deploying your models letting you put your machine learning projects into action smoothly under computer vision

workloads we have image classification object detection optical character recognition facial detection and facial

analysis Solutions next we have Azure AI Vision Azure AI face detection and Azure AI

video indexer the Azure AI Services Encompass a wide range of tools designed to facilitate the development of

intelligent applications these Services used to be called computer vision custom vision face service and form recognizer

but have Evol or been grouped under broader service categories to streamline their application and integration into

projects for NLP we have key phrase extraction entity recognition sentiment analysis language modeling speech

recognition synthesis this one doesn't really appear much it's kind of a concept not so much something we have to

do and then there's translation so now we have Azure tools and services for NLP workloads these

include the Azure AI language service Azure AI speech service and Azure AI translator service these used to be

separate Services I believe like the text analytics service Lewis speech service and translator text service but

they have been added to the Azure AI umbrella of AI services and now we'll be moving on to the generative AI workloads

on Azure we'll be covering features of generative AI models common scenarios for generative Ai and responsible AI

considerations for generative Ai and also some of the cool features that Azure open ey service has to offer such

as natural language generation code generation and image generation so that's about a general

breakdown of the AI 900 exam [Music] guide hey this is angrew Brown from exam

Pro and we are looking at the layers of machine learning so here I have this thing that looks like kind of an onion

and what it is it's just describing the relationship between these uh ml terms uh uh related to Ai and we'll just work

our way through here starting at the top so artificial intelligence also known as AI is when machines that perform jobs

that mimic human behavior so it doesn't describe uh how it does that but it's just the fact that that's what AI is uh

one layer underneath we have machine learning so machines that get better at a task without explicit programming uh

then we have deep learning so these are machines that have an artificial neural network inspired by the human brain to

solve complex problems and if you're talking about someone that actually assembles either ml or or deep learning

uh models or algorithms that's a data scientist so a person with multi-disciplinary skills and math

statistics predictive modeling machine learning to make future predictions so what you need to understand is that AI

is just the outcome right and so AI could be using ml underneath or deep learning or a combination of both or

just IFL statements okay [Music] all right so let's take a look here at

the key elements of AI so AI is the software that imitates human behaviors and capabilities and there are key

elements according to Azure or Microsoft as to what makes up AI so let's go through this list quickly here so we

have machine learning which is the foundation of an AI system that can learn and predict like a human you have

anomaly detection so detect outliers or things out of place like a human computer vision be able to see like a

human natural language processing also known as NLP be able to process human languages in referr contexts you know

like a human at conversational AI be able to hold a conversation with a human so you know I wrote here according to

Microsoft and Azure because you know the the global definition is a bit different but I just wanted to put this here

because I've definitely seen this as an exam question and so we're going to have to go with azure's definition here okay

let's define what is a data set so a data set is a logical grouping of units of data that are closely related to or

share the same data structure and there are publicly available data sets that are used in uh learning of Statistics

data analytics and machine learning I just want to cover a couple here so the first is the mnist database so images of

handwritten digits used to test classify cluster image processing algorithms commonly used when learning uh how to

build computer vision ml models to translate handwritten into or handwriting into digital text so it's

just a bunch of handwritten uh numbers and letters and then another very popular data set is the common objects

in context Coco data set so this is a data set which contains many common images using a Json file Coco format

that identify objects or segments within an image uh and so this data set has a lot of stuff in it so object

segmentations recognition and it contexts super pixel stuff segmentation they have a lot of images and a lot of

objects uh so there's a lot of stuff in there so why am I talking about this and in particular Coco data sets well when

you use um Azure machine Learning Studio it has a DAT data labeling service and um the thing is is that uh it can

actually export out into Coco formats that's why I wanted you to get exposure to what Coco was and the other thing is

is that when you're building out Azure machine learning uh pipelines you uh they actually have open data sets which

we'll see later in the course um that shows you that you can just use very common ones and so uh you might see mest

and uh the other one there uh so I just wanted to get you some exposure [Music]

okay let's talk about data labeling so this is the process of identifying raw data so images text files videos and

adding one or more meaningful and informative labels to provide context so a machine learning model can learn so

with supervised machine learning labeling is a prerequisite to produce training data and each piece of data

will generally be labeled by a human the reason why I say generally here is because with azure's uh data labeling

Service uh they can actually do ml assisted labeling uh so with unsupervised machine learning labels

will be produced by the machine and may not be human readable uh and then one other thing I want to touch on is the

term called Ground truth so this is a proper u a properly labeled data set that you can use as the objective

standard to train and assess a given model is often called Ground truth the accuracy of your train model will depend

on the accuracy of your ground Truth Now using um azures tools I never seen use the word ground truth I see that a lot

in AWS and even this graphing here is from AWS but uh I just want to make sure you are familiar with all that stuff

[Music] okay let's compare supervised unsupervised and reinforcement learning

starting at the top we got supervised learning this is where the data has been labeled for training and it's considered

Tas driven because you are trying to make a prediction get a value back so when the labels are known and you want a

precise outcome when you need a specific value returned and so you're going to be using classification and regression in

these cases for unsupervised learning this is where data that has not been labeled uh the ml model needs to do its

own labeling this is considered data driven it's trying to recognize a structure or a pattern and so this is

when the labels are not known and the outcome does not need to be prise when you're trying to make sense of data so

you have clustering dimensionality reduction and associ if You' never heard this term before the

idea is it's trying to reduce the amount of Dimensions to make it easier to work with the data so make sense of the data

right uh we have reinforcement learning so this is where there is no data there's an environment and an ml model

generates data uh and and makes many attempts to reach a goal so this is considered uh decisions driven and so

this is for game AI learning tasks robot navigation when you've seen someone code a video game that can play itself that's

what this is if you're wondering this is not all the types of machine learning uh and these are in specific unsupervised

and supervised is considered classical machine learning because they he heavily rely on statistics and math to produce

the outcome uh but there you [Music] go so what is a neural network well it's

often described as mimicking the brain it's a neuron or node that represents an algorithm so data is inputed into a

neuron and based on the output the data will be passed to one of many connected neurals the connections between neurons

is weighted I really should have highlighted that one that's very important uh the network is organized

into layers there will be an input layer uh one to many hidden layers and an output layer so here's an example of a

very simple neural network notice the NN a lot of times you'll see this in ml as an abbreviation for neural networks and

sometimes neural networks are just called neural Nets so just understand that's the same term here what is deep

learning this is a neural network that has three or more hidden layers it's considered deep learning because at this

point it's uh it's not human readable to understand what's going on with within those

layers what is forward feed so neural networks where they have connections between nodes that do not form a cycle

they always move forward so that just describes uh a a forward pass through the network you'll see fnn which stands

for forward feed neural network just to describe that type of network uh then there's back back propagation which are

in forward feed uh networks this is where we move backwards through the neural net adjusting the weights to

improve the outcome on next iteration this is how a neural net learns the way the back propagation knows to do this is

that there's a loss function so a function that compares the ground truth to the prediction to determine the error

rate how bad the network performs so when it gets to the end it's going to perform that calculation and then it's

going to do its back propagation and adjust the weights um then you have activation functions I'm just going to

uh clear this up here so activation functions uh they're an algorithm applied to a hidden layer uh node that

affects connected output so for this entire hidden layer they'll all have the same uh one here and it just kind of

affects uh how it learns and like how the waiting works so it's part of back propagation and just the learning

process there's a concept of D so when the next layer increases the amount of nodes and you have spars so when the

next layer decreases the amount of nodes anytime you see something going from a dense layer to a sparse layer that's

usually called dimensional dimensionality reduction because you're reducing the amount of Dimensions

because the amount of nodes in your network determines the dimensions you have

okay [Music] what is a GPU well it's a general

processing unit that is specially designed to quickly uh render high resolution images and videos

concurrently gpus can perform parallel operations on multiple sets of data so they are commonly used for non-graphical

tasks such as machine learning and scientific computation so a CPU has an average of four to 16 processor cores a

GPU can have thousands of processor cores so something that has 408 gpus could have as many as 40,000 cores

here's an image I grabbed right off the Nvidia website and so it really illustrates very well uh like how this

would be really good for machine learning or U neural networks because no networks have a bunch of nodes they're

very repetitive tasks if you can spread them across a lot of cores that's going to work out really great so gpus are

suited uh for repetitive and highly paralleled Computing tasks such as rendering Graphics cryptocurrency mining

deep learning and machine [Music] learning we're talking about Cuda before

we can let's talk about what Nvidia is so Nvidia is a company that manufactures graphical processor units for gaming and

professional markets if you play video games you've heard of Nvidia so what is Cuda it is the compute unified device

architecture it is a parallel Computing platform in API by Nvidia that allows developers to use Cuda enable gpus for

general purpose Computing on gpus so gpg all major deep learning Frameworks are integrated with Nvidia deep uh

learning SDK the Nvidia uh deep learning SDK is a collection of Nvidia libraries for deep learning one of those libraries

is the Cuda deep neural network library so CNN so Cuda or CNN provides highly tuned implementations for standard

routines such as forward and back uh convolution convolution is really great for um uh uh computer vision pooling

normalization activation layers uh so you know in the Azure certification uh for the AI 900 uh they're not going to

be talking about Cuda but if you understand these two things you'll understand why gpus uh really matter

[Music] okay all right let's get a uh easy introduction into machine learning

pipeline so this one is definitely not an exhaustive one and we're definitely going to see more complex ones uh

throughout this course but let's get to it here so starting on the left hand side we might start with data labeling

this is very uh important when you're doing supervis learning because you need to label your data so the ml model can

learn by example during training uh this stage and the feature engineering stage are is considered pre-processing because

we are preparing our data to be trained for the model uh when we move on to feature engineering the idea here is

that ml models can only work with numerical data so you'll need to translate it into a format that it can

understand so extract out the important data that the ml model needs to focus on okay uh then there's the training steps

so your model needs to learn how to become smarter it will perform multiple iterations getting smarter with each

iteration uh you might also have a hyperparameter tuning uh step here it says tunning but

it should say tuning um but the ml model can have different parameters so you can use ml to try out many different

parameters to optimize the outcome when you get to deep learning it's impossible to tweak the parameters by hand so you

have to use hyperparameter tuning then you have serving sometimes known as deploying uh but you know when we say

deploy we talk about the entire pipeline not necessarily just the the ml model step so we need to make an ml model

accessible so we serve it by hosting in a virtual machine or container uh when we're talking about Azure um machine

learning it's either going to be an Azure kubernetes service or Azure container instance and you have uh

inference so inference is the active request uh of requesting to make a prediction so you send your payload with

either CSV or whatever and you get back the results you have a real time endpoint and batch processing so real

time is just they can batch can be real as well but generally it's slower but the idea is that do I am I making a

single item prediction or am I giving you a bunch of data at once and again this is a very simplified ml pipeline

I'm sure we'll revisit ml pipeline later in this [Music]

course so let's compare the uh the terms forecasting and prediction so forecasting you make a prediction with

relevant data it's great for analysis of Trends uh and it's not guessing and when you're talking about prediction this is

where you make a prediction without relevant data you use statistics to predict future outcomes it's more of

guessing and it uses decision Theory so imagine you have a bunch of data and the idea is you're going to infer from that

data okay maybe it's a maybe it's B maybe it's C and for prediction you don't have really much data so you're

going to have to uh kind of invent it and the idea is that you'll figure out what the outcome is there these are

extremely broad terms but you know just so you have a highle view of these two things

okay so what are performance or evaluation metrics well they are used to evaluate

different machine learning algorithms the idea is uh you know when your machine learning makes a prediction

these are the metrics you're using to evaluate to determine you know is your ml model working as you intended so for

different types of problems different metrics matter this is absolutely not an exhaustive list I just want you to get

you exposure to these uh words and things uh so that when you see them you go okay I'll come back here and refer to

this uh but lots of these you're just it's not it's not necessarily to remember but classification metrics you

should know so classification we have accuracy precision recall F1 score rock and a for regression metrics we have MSE

R msce Mae ranking metrics we have MMR dcg and dcg statistical metrics we have correlation computer vision metrics we

have psnr ssim IOU NLP metrics we have perplexity blue medor Rogue deep learning related metrics we have

Inception score I cannot say this person's name but I'm assuming it's a person but uh this Inception distance

and there are two categories of evaluation metrics we have internal evaluations so metrics used to evaluate

the internals of an ml model so accuracy F1 score Precision recall I call them the famous four using all kinds of uh

models and uh external evaluation metrics used to evaluate the final prediction of an ml model so yeah uh

don't get too worked up here I know that's a lot of stuff uh the ones that matter we will see again again

[Music] okay let's take a look at Jupiter notebook so these are web-based

applications for authoring documents that combin live code narrative text equations visualizations uh so if you're

doing data science or you're building ml models you absolutely are going to be working with jupyter notebooks they're

always integrated into uh cloud service providers ml tools um uh so jupyter notebook actually came about from

IPython so IPython is the precursor of it and they extracted that feature out it became jupyter notebook I IPython is

now a kernel uh to run uh python so when you execute out python code here it's using IPython which is just a version of

python uh jupyter notebooks were overhauled and better integrated into an IDE called Jupiter Labs which we'll talk

about here in a moment and you generally want to open notebooks in Labs the Legacy webbased interface is known as

Jupiter classic notebooks so this is what the old one looks like you can still open them up but everyone uses

Jupiter Labs now okay so let's talk about Jupiter labs jupyter labs is the next generation web-based user interface

all familiar features of the classic Jupiter notebook uh is in a flexible powerful user interface it has notebooks

a terminal a text editor a file browser Rich outputs Jupiter Labs will eventually replace the classic uh

Jupiter notebooks so there you [Music] go we keep mentioning regression but

let's talk about it in uh more detail here so we kind of understand the concept so regression is the process of

finding a function to equate a labeled data set notice it says labeled that means it's going to be for supervised

learning into a continuous variable number so another way to say it is predict this variable in the future so

the future is just means like that continuous variable doesn't have to be time but that's just a good example of

regression so what will the temperature be next week so will it be 20 Celsius how would we determine that well we

would have vectors so dots that are plotted on a graph that has multiple Dimensions the dimensions could be

greater than just X and Y you could have uh many uh and then you have a regression Line

This is the line that's going through our data set and uh and that's going to help us uh figure out um how to predict

the value so how would we do that well we would need to calculate the distance of a vector from the regression line

which is called an error and so different regression algorithms use uh the error to predict different variable

future variables so just to look at this graphic here so here is our regression line and here is a a DOT like a a vector

a piece of information and this distance from the line the the actual distance is what we're going to use in our ml model

to figure out if we were to plot another line up here right you know we would compare this line to all the other lines

okay and that's how we'd find similarity and what we'll commonly see for this is mean squared error root mean squared

error mean absolute error so MSE mrse and Mae [Music]

okay let's take a closer look at the concepts of classification so classification is the process of finding

a function to divide a labeled data set so again this is supervised learning into classes or categories so predict a

category to apply to the inputed data so will it rain next Saturday will it be sunny or rainy so we have our data set

and the idea is we're drawing through this a classification line to divide the data set so we're regression we're

measuring the line two or the the the vectors to the line and this one it's just what side of the line is it on if

it's on this side then it's sunny if it's on this side it's rainy okay for classification algorithms we got log

logistic regression decision trees random forests neural networks uh naive Bays K nearest neighbor also known as

knnn and support Vector machines svms [Music] okay let's take a closer look at

clustering so clustering is the process of grouping unlabeled data so unlabeled data means it's unsupervised learning

based on similarity and differences so the outcome could be group data based on similarities or differences I guess it's

the same description up here uh but imagine we have a graph and we have data and the idea is we draw boundaries

around that to see uh similar groups so maybe we're recommending purchases to Windows computers or recommending

purchase to Mac computers now remember this is unlabeled data so the label is being inferred or um or they're just

saying these things are similar right so clustering algorithms we got K means k medoids a density Bas hierarchial

[Music] okay hey this is Andrew Brown from exam Pro and we're looking at the confusing

Matrix and this is a table to visualize the model predictions the predicted versus the ground truth labels the

actual also known as an error Matrix and they are useful for classification problems to determine if our um if our

classification is working as we think it is so imagine we have a question how many bananas did this person eat or

these people eat and so we have this kind of a box here where we have predicted versus actual and it's really

comparing the ground truth and what the model predicted right and so on the exam they'll ask you questions like okay well

imagine that uh and they might not even say yes or no maybe like zero and one and so what they're saying is you know

imagine you have you want to tell us the true positives right and so the idea is they won't show you the labels here but

you know one and one would be a true positive and zero and Z would be a false negative okay another thing they'll ask

you about these uh confusion matrixes is uh the size of them so the idea is that we're looking right now at a um oops

just going to erase that there but we are looking at a binary classifier because we have one label and uh uh just

two labels right one and two okay but you could have three say one two and three so how would you calculate that

well there would just be a third cell over here uh you know and it's just going to be actual predicted because

we're only going to have ground Truth Versus prediction and so that's how you'll know it will be six the size will

be six might not say cells but it'll just say six [Music]

okay so to understand anomaly detection let's define quickly what is an anomaly so an abnormal thing that is marked by

deviation from the norm or standard so anomaly detection is the process of finding outliers within a data set

called an anomaly so detecting when a piece of data or access pattern appear suspicious or malicious so use

cases for Nom detection can be data cleaning intrusion detection fraud detection system Health monitoring event

detection and sensory or sensor networks ecosystem disturbances detection of critical and cascading flaws Anomaly

detections by hand is a very tedious process so using ml for anomaly detection is more efficient and accurate

and Azure has a service called anomaly detector detects anomalies and data to quickly find uh quick identify and

troubleshoot [Music] issues so computer vision is when we use

machine learning neural networks to gain high level understanding of digital images or videos so for computer vision

deep learning algorithms we have convolutional neural networks these are for image and video recognition they're

inspired after how the human eye actually processes information and sends it back to the brain to be processed you

have recurrent neural networks rnns which are generally used for handri recognition or speech recognition of

course these algorithms have other applications but these are the most common use cases for them for types of

computer vision we have image classification so look at an image or video and classify its place in a

category object detection so identify objects within an image or video and apply labels and location boundaries

semantic segmentation so identify segments or objects by drawing pixel masks around them so great for objects

and movement image analysis so analyze uh an image or video to apply descriptive context uh labels so maybe

an employee is sitting at a desk in Tokyo would be uh something that image analysis would do optical character

recognition or OCR find texts in images or videos and extract them into digital text for editing facial detection so

detect faces in a photo or video and draw a location boundary uh and label their expression so for computer vision

to some things around Azure or Microsoft Services there's one called seeing AI it's an AI app developed by Microsoft

for iOS and so you use your device camera to identify OB uh people and objects and the app is audibly describes

those objects for people with visual impairments it's totally free if you have an IOS app I have an Android phone

so I cannot use it but I hear it's great some of the Azure computer vision service offering is computer vision so

analyze images and videos extract descriptions tags objects and texts custom Vision so custom uh image

classification object detection models using your own images face so detect and identify people uh emotions and images

form recognizer so translate scan documents into key value or tabular editable

[Music] data so natural language processing also known as NLP is machine learning that

can understand the context of a corpus Corpus being a body of related text so nlps enable you to analyze and interpret

text within documents and email messages interpret or contextualize spoken tokens so for example maybe customer sentiment

analysis whether customer is happy or sad synthesize speech so a voice assistant uh assistant talking to you

automatically translates spoken or written phrases and sentences between languages interpret spoken or written

commands and determine appropriate actions a very famous example for a voice assistant specifically or virtual

assistant for Microsoft is Cortana uh it uses the Bing search engine to perform tasks such as setting reminders and

answering questions for the user uh and if you're on a Windows 10 machine uh it's very easy to activate Cortana by

accident uh when we're talking about azures MLP offering we have text and analytics so sentiment analysis to find

out what customers think find topic uh topic relevant phrases using key phrase extraction identify the language of the

text with the language detection detect and categorize entities in your text with named entity recognition for

translator we have real-time text translation multilanguage support uh for speech service we have transcribe

audible speech into readable searchable texts and then we have language understanding also known as

Lewis uh natural language processing service that enables you to understand human

language in your own application website chatbots iot device and more when we talk about conversational AI it usually

generally uses NLP so that's where you'll see that overlap next [Music]

okay let's take a look here at conversational AI which is technology that can participate in conversations

with humans so we have chat Bots voice assistants and interactive voice recognition systems which is like the

the second version to interactive voice response system so you know when you call in and they say press these numbers

that is a response system and a recognition system is when they can actually take human uh Speech and

translate that into action so the use cases here would be online customer support replaces human agents for uh for

replying about customer FAQs maybe shipping questions anything about customer support accessibility so voice

operate UI for those who are uh visually impaired HR processes so employee training onboarding updating employee

information I've never seen it used like that but that's what they say as a use case Healthcare accessible affordable

healthare so maybe you're doing a claim process I've never seen this but maybe in the US where you do more your claims

and everything is privatized it makes more sense Internet of Things So iot devices so Amazon Alexa Apple Siri

Google home and I suppose Cortana but it doesn't really have a particular device so that's why I didn't list it there

computer software so autocomplete search on phone or desktop so that would be Cortana something it could do uh for the

two services that are around conversation AI for Azure we have Q&A maker so create a conversational

question and answer bot from your existing content also known as a knowledge base and Azure bot service

intelligent serverless bot service that scales on demand used for creating publishing managing bots so uh the idea

is you make your Bot here and then you deploy it with this [Music]

okay let's take a look here at responsible AI which focuses on ethical transparent and accountable uses of AI

technology Microsoft put into practice responsible AI via its six Microsoft AI principles this whole thing is invented

by Microsoft uh and so you know it's not necessarily a standard but it's something that Microsoft is pushing hard

to uh have people adopt okay so we the first thing we have is fairness so this is an AI system which should treat all

people fairly we have reliability and safety an AI system should perform reliably and safely privacy and security

AI system should be secure and respect privacy inclusiveness AI system should Empower everyone and engage people

transparency AI systems should be understandable accountability people should be accountable for AI systems and

we need to know these in uh uh greater detail so we're going to have a a short little video on each of these

[Music] okay the first on our list is fairness so AI systems should treat all people

fairly so an AI system can reinforce existing social social stere uh stereotypical bias can be introduced uh

during the development of a pipeline so an A system that are used to allocate or withhold opportunities resources or

information uh in domains such as criminal justice employee employment and hiring finance and credit so an example

here would be an ml model designed to select a final applicant for hiring pipeline without incorporating any bias

based on gender ethnicity or may result in unfair Advantage so Azure ml can tell you how each feature can influence a

model's prediction for bias uh one thing that could be of use is fair learn so it's an open source python project to

help data science is to improve fairness in the AI systems at the time of I made this course a lot of their stuff is

still in preview so you know it's the fairness component is is not 100% there but it's great to see that they're

getting that along okay so we are on to our second AI principle for Microsoft and this one is

AI systems should perform reliably and safely so AI software must be rigorously tested to ensure they work as expected

before release to the end user if there are scenarios where AI is making mistakes it is important to release a

report Quantified risks and harms to end users so they are informed of the shortcomings of an AI solution something

you should really remember for the exam they'll definitely ask that AI wear concern uh for reliability safety for

humans is critically important autonomous vehicles a health diagnosis a suggestion prescriptions and autonomous

weapon systems they didn't mention this in their content and I was just like doing some additional research research

I'm like yeah you really don't want mistakes when you have automated weapons or ethnically you shouldn't have them at

all but hey that's uh that's just how the world works but yeah this is this category

[Music] here we're on to our third Microsoft AI principle AI system should be secure and

respect privacy so AI can require vast amounts of data to train deep machine ml models the nature of an ml model may

require personally identifiable information so piis uh it is important that we ensure

protection of user data that it is not leak or disclosed in some cases ml models can be run locally on a user's

device so their uh piis remain on their device avoiding the the vulnerability this is called this is like Edge

Computing so that's the concept there AI security principles to malicious actors so data origin and lineage data use

internal versus external data corruption considerations anomaly detection so there you

[Music] go we're on to the fourth Microsoft AI principle so AI systems should Empower

everyone and engage people if we can design AI Solutions for the minority of users they can design a solution for the

majority of users so when we're talking about minority groups we're talking about physical ability gender sexual

orientation ethnicity other factors this one's really simple uh in terms of practicality it doesn't 100% make sense

because if you've worked with um uh groups that are deaf and blind developing technology for them a lot of

times they need specialized Solutions uh but the approach here is that you know if we can design for the minority we can

design for all that is uh the principle there so that's what we need to know okay

let's take a look here at transparency so AI systems should be understandable so interpretability and intelligibility

is when the end user can understand the behavior of UI so transparency of AI systems can result in mitigating

unfairness help developers debug their AI systems ging more trust from our users those build a those who build AI

systems should be open about why they're using AI open about the limitations of the AI systems adopting an open source

AI framework can provide transparency at least from a technical perspective on the internal workings of an AI

system we are on to the last Microsoft AI principle here people uh should be accountable for AI systems so the

structure put in place to consistently enacting AI principles and taking them into account AI systems should work

within Frameworks of governments organizational principles ethical and legal standards that are clearly defined

principles guide Microsoft and how they develop sell and Advocate when working with third parties and this push towards

regulation towards a principles so this is Microsoft saying hey everybody adopt our model um there aren't many other

model so I guess it's great that Microsoft is taking the charge there I just feel that it needs to be a bit more

welldeveloped but what we'll do is look at some more practical examples so we can better understand how to apply their

principles okay so if we really want to understand how to apply the Microsoft AI principles

they've great created this nice little tool via a free web app for practical scenarios so they have these cards you

can read through these cards they're color coded for different scenarios and there's a website so let's go take a

look at that and see what we can learn [Music] okay all right so we're here on the

guidelines for human AI interaction so we can better understand the uh how to put into practice the Microsoft AI

principles they have 18 cards and let's work our way through here and see the examples the first one our list make

clear what the system can do help the users understand what the AI system is capable of doing so here PowerPoint

quick start Builders an on uh Builds an online outline to help you get started researching a subject it display uh

suggested topics that help you understand the features capability then we have the Bing app shows examples of

types of things you can search for um Apple watch displays all metrics it tracks and explains how going on the

second card we have make clear how well the system can do what it can do so here we have office new uh companion

experience ideas dock alongside your work work and offers one-click assistance with grammar design Data

Insights richer images and more the unassuming term ideas coupled with label previews help set expectations and

presented suggestions the recommender in apple music uses language such as we will think you'll like to communicate

uncertainty the help page for Outlook web mail explains the filtering into focused and other and we'll start

working right away but we'll get better with use making clear the mistakes uh will happen and you teach the product

and set overrides onto our red cards here we have time Services based on context time when to act or interrupt

based on the user's current task and environment when it's time to leave for appointments Outlook sends a time to

leave notification with directions for both driving and public transit taking into account current location event

location real-time traffic information um and then we have after using Apple Maps routing it remembers

when you're parked your car when you open the app after for a little while it suggests routing to the location of the

park car all these Apple examples make me think that Microsoft has some kind of partnership with apple I guess I guess

Microsoft or or Bill Gates did own Apple shares so maybe they're closer than we think uh show contextually relevant

information time when to act or interrupt based on user's current task and environment powered by Machine

learning acronyms in word helps you understand shorthand employed uh in your own work environment relative to current

Open document uh on Walmart.com when the user is looking at a product such as gaming

console recommends accessories and games that would go with it when a user searches for movies Google shows results

including showtimes near the users's location for the current data onto our fifth card here match based uh we didn't

we didn't miss this one right yeah we did okay so we're on the fifth one here match relevant social norms ensure

experience is delivered in a way the users would expect given the social cultural context when editor identifies

ways to improve writing style presents options politely consider using that's the Canadian way being polite uh Google

photos is able to recognize pets and use the wording important cats and dogs recognizing that for many pets are an

important part of one's family and you know what uh when I uh started renting my new house uh I I said you know is

there a problem with dogs and my landlord said well of course pets are part of the family and that was

something I like to hear uh Cortana uses semiformal tone apologizing when unable to find a uh contact which is polite and

socially appropriate I like that okay mitigate social biases ensure AI system languages and behaviors do not

reinforce undesirable unfair stereotypes and biases my anal summarizes how you spend your time at work then suggest

ways to work smarter one ways to mitigate bias is by using gender neutral icons to represent important people

sounds good to me a Bing search for CEO or doctor shows images of diverse people in terms

of gender and an ethnicity sounds good to me the predictive uh keyboard for Android suggests both genders when

typing a pronoun starting with the letter H we're on to our yellow cards uh so support efficient invocation so make

it easy to invoke or request system Services when needed so flashfill is a helpful timesaver in Excel that can be

easily invoked with on canvas interactions and uh that keep you in flow on amazon.com oh hey there got

Amazon in addition to the system giving recommendations as you browse you can manually invoke additional

recommendations from the recommender for your menu uh design ideas in Microsoft PowerPoint can be invoked uh with the

with the Press of a button if needed I cannot stand it when that pops up I always have to tell it to leave me alone

okay support efficient dismal uh efficient disle dismissal oh support efficient

dismissal okay make it easy to dismiss or ignore or undesired AI system Services okay this sounds good to me

Microsoft forms allows you to create custom surveys quizzes polls questionnaires and forms some choices

questions trigger suggested options position beneath the relevant question the suggestion can be easily ignored and

dismissed Instagram allows the user to easily hide or report ads that have been suggested by AI by tapping the ellipses

at the top of the right of the ad Siri can be easily dismissed uh uh by saying never mind

I'm always telling my Alexa never mind support efficient uh correction make it easy to edit refine or recover the AI

system uh when the when the AI system is wrong so alt Auto alt text automatically generates alt text for photographs by

using intelligent services in the cloud descriptions can be easily Modified by clicking the alt text button in the

ribbon once you set a reminder with Siri the UI displays a tap to edit link when Bing automatically corrects spelling

errors in search queries it provides the option to revert to the query as originally typed with one click on to

card number 10 Scope Services when in doubt so engage in

disambiguate disambiguation or gracefully degrade the AI system service when uncertain about a

user's goal so when Auto replacing word is uncertain of a correction it engages in disambiguation by displaying multiple

options you can select from Siri will let you know it has trouble hearing if you don't respond or talk or or speak

too softly big Maps will provide multiple routing options when uh when unable to recommend best one we're on to

card number 11 make clear why the system did what it did enable users to access an explanation of why the AI system

behaved as it did office online recommends docu documents based on history and activity descriptive text

above each document makes it clear why the recommendation is shown product recommendations on Amazon .c include why

Rec recommended link that shows that what products in the user shopping history informs the recommendations

Facebook enables you to access an explanation about why you are seeing each ad in the news

feed onto our green cards so remember recent interactions so maintain short-term memory and allow the user to

make efficient references to that memory when attaching a file Outlook offers a list of recent files including recently

copied file links Outlook also remembers people you have interacted with recently and displays uh them when addressing a

new email uh Bing search remembers some recent queries and search can be

continued uh conversationally how old is he after a search for kyanu Reeves Siri carries over the context from one

interaction to the next a text message is Created from the person you told Siri to message to onto card number 13 lucky

number 13 learn from user Behavior personalize the user experience by learning from their actions over time

tap on a search bar in office applications and search lists uh the top three commands on your screen that

you're most likely to need to personalize the technology called zero query doesn't even need to type in the

search bar to provide a personalized predictive answer amazon.com gives personalized product recomm

recommendations based on previous purchases onto card 14 update and adapt Cur uh C uh cautiously limit disruptive

changes when updating adaptive adapting the systems behaviors so PowerPoint designer improves slides for office 65

subscribers by automatically generating design ideas from to choose from designer has integrated new capabilities

such as smart Graphics icon suggestions and existing user experience ensuring the updates are not

disruptive office tell office tell me feature shows dynamically recommended items and a designated try area to

minimize disruptive changes onto card number 15 encourage granular feedback back enable the users to provide

feedback indicating their preferences during regular interactions with the AI system so ideas and Excel empowers you

to understand your data through high level visual summaries Trends and patterns encourages feedback on each

suggestion by asking is this helpful not only does Instagram provide the option to hide specific ads but it also

solicits feedback to understand why the ad is not relevant and Apple's music app love dislike buttons are prominent

easily accessible number 16 convey the consequences of us actions immediately

update or convey how user actions will impact future behaviors of the AI system you can get stock in G Geographic data

types and Excel it is easy as typing text into a cell and converting it to stock data type or geograph geography

data type when you perform the conversion action an icon immediately appears in the converted cells uh upon

tapping the like dislike button for each recommendation it at in apple music a pop-up informs the user that they'll

receive more or fewer similar recommendations onto card number 17 we're almost near the end provide Global

controls allow the user to globally customize the system system monitors and how it behaves so editor expands on

spelling and grammar checking capabilities of word to include more advanced proofing and editing designed

to ensure document is readable editor can flag a range of critique types and allow to customize the thing is is that

in word it's so awful spellchecking I don't understand like it's been years and the the spell checking never gets

better so they got to emplore better spellchecking AI I think bang search provides settings that impact the the

types of results the engine will return for example safe search uh then we have Google photos

allows user to turn location history on enough for future photos it's kind of funny seeing like Bing in there about

like using AI because at one point it's almost pretty certain that Bing was copying just Google search indexes to

learn how to index I don't know that's Microsoft for you uh we're on on to card 18 notify users about changes inform the

user when AI system adds or updates his capabilities uh the what's new dialogue in office informs you about changes by

giving an overview of the latest features and updates including updates to AI features in Outlook web the help

tab includes a what's new section that covers updates so there we go we made it to the end of the list I hope that was a

fun listen for you and and there I hope that we could kind of match up the uh the responsible AI I kind of wish what

they would have done is actually mapped it out here and said where it match but I guess it's kind of an isolate service

that kind of ties in so I guess there we go [Music]

okay hey this is Andrew Brown from exam Pro and we are looking at Azure cognitive services and this is a

comprehensive family of AI services and cognitive apis to help you build intelligent apps so create customizable

pre-trained models built with breakthrough AI researches I put that in quotations I'm kind of throwing some

shade at uh Microsoft at Azure just because it's their marketing material right uh deploy cognitive Services

anywhere from Cloud to the edge uh with containers get started quickly no machine learning expertise required but

I think it it helps to have a bit of background knowledge uh develop with strict ethical standards uh Microsoft

loves talking about the responsible um their responsible AI stuff empowering responsible use with industry leading

tools and guidelines so let's do a quick breakdown of the types of services in this family so for decision we have

anomaly detector identify potential problems early on content moderator detect potentially offensive or unwanted

content personalizer create Rich personalized experiences for every user for languages we have language

understanding also known as uh Luis Lewis I don't know why I didn't put the initialism there but don't worry we'll

see it again build natural language understanding into app spots and iot devices Q&A maker create a