Slide 1

Slide 1 text

ML Starter Pack A comprehensive guide to Machine Learning April 22, 2023 GDG Waterloo, Canada

Slide 2

Slide 2 text

I'm Charmi! ● ML Grad at Mila and UdeM ● GDE for ML ● International Speaker ● Worked @ AWS, Shipmnts, ISRO ● Tech, Art, Travel

Slide 3

Slide 3 text

Plan of the Day! ● What is Machine Learning? ● Types of ML ● Supervised and Unsupervised techniques ● Classification and Regression ● Basics of Deep Learning ● Basics of NLP, CV, ASR ● The 3D ML Pipeline ● Fun Applications

Slide 4

Slide 4 text

? 90% have failed to solve this... = = = 33 30 36

Slide 5

Slide 5 text

99% have failed to solve this...

Slide 6

Slide 6 text

Retrospection time!! What we just did? We solved logical math puzzle in a fraction of time

Slide 7

Slide 7 text

Think about, How we solved them? We had inputs We also had outputs

Slide 8

Slide 8 text

Think about, How we solved them? We created Rules and Patterns in our brain!!

Slide 9

Slide 9 text

Can machines create rules on their own?

Slide 10

Slide 10 text

Can machines create rules on their own? No, if it’s Classical Programming...

Slide 11

Slide 11 text

Can machines create rules on their own? YES, if it’s Machine Learning!!!

Slide 12

Slide 12 text

Classical Programming vs ML Classical Programming Machine Learning Rules Rules Data Data Answers Answers

Slide 13

Slide 13 text

Artificial Intelligence Deep Learning Machine Learning Any Technique that enables computers to mimic human intelligence & behaviour A subset of ML exposing multilayered neural networks to vast amount of data A subset of AI that includes statistical techniques to solve the tasks using experience

Slide 14

Slide 14 text

No content

Slide 15

Slide 15 text

Basic Terminologies ● Features ● Labels ● Examples ○ Labelled example ○ Unlabelled example ● Data Split (Train, Valid, Test) ● Models (Train and Test) ○ Classification model ○ Regression model

Slide 16

Slide 16 text

Supervised Learning ● Supervised Learning deals with prediction of values based on given combinations of data and labels given beforehand ● ML systems learn how to combine input to produce useful predictions on never-before-seen data ● It is like learning with a teacher

Slide 17

Slide 17 text

Regression and Classification ● A regression model predicts continuous values by fitting a line ○ What is the value of a house in California? ○ What is the probability that a user will click on this ad? ● A classification model predicts discrete values by creating boundaries ○ Is a given email message spam or not spam? ○ Is this an image of a dog, a cat, or a hamster?

Slide 18

Slide 18 text

Iterative Learning

Slide 19

Slide 19 text

Overfitting vs Underfitting ● An overfit model gets a low loss during training but does a poor job predicting new data ● Overfitting is caused by making a model more complex than necessary ● The fundamental tension of machine learning is between fitting our data well, but also fitting the data as simply as possible

Slide 20

Slide 20 text

Unsupervised Learning ● It deals with clustering values or forming groups of values ● One aims to infer patterns from the data rather than predicting values ● It is like learning on your own

Slide 21

Slide 21 text

Evaluation Metrics https://en.wikipedia.org/wiki/Precision_and_recall, https://www.kdnuggets.com/2018/06/right-metric-evaluating-machine-learning-models-2.html

Slide 22

Slide 22 text

Evaluation Metrics https://en.wikipedia.org/wiki/Precision_and_recall, https://www.kdnuggets.com/2018/06/right-metric-evaluating-machine-learning-models-2.html

Slide 23

Slide 23 text

No content

Slide 24

Slide 24 text

Deep Learning

Slide 25

Slide 25 text

Neural Network https://gfycat.com/gifs/search/neural+network

Slide 26

Slide 26 text

When to use or not use DL? ● Deep Learning outperforms other techniques if the data size is large. But with small data size, traditional Machine Learning algorithms are preferable ● Finding large amount of “Good” data is always a painful task ● Deep Learning techniques need to have high end infrastructure to train in reasonable time ● When there is lack of domain understanding for feature introspection, Deep Learning techniques outshines others as you have to worry less about feature engineering

Slide 27

Slide 27 text

When to use or not use DL? ● Model Training time: a Deep Learning algorithm may take weeks or months whereas, traditional Machine Learning algorithms take few seconds or hours ● DL never reveals the “how and why” behind the output- it’s a Black Box ● Deep Learning really shines when it comes to complex problems such as image classification, natural language processing, and speech recognition ● DL excels in tasks where the basic unit (pixel, word) has very little meaning in itself, but the combination of such units has a useful meaning

Slide 28

Slide 28 text

Natural Language Processing ● Sentiment analysis ● Chatbots ● Machine translation ● Speech recognition ● Text summarization

Slide 29

Slide 29 text

Computer Vision ● Facial recognition ● Object detection ● Medical imaging ● Autonomous vehicles ● Segmentation

Slide 30

Slide 30 text

Speech Understanding ● Voice-controlled personal assistants ● Virtual agents ● Dictation software ● Closed captioning ● Subtitling for movies ● Transcribing

Slide 31

Slide 31 text

The 3D ML Pipeline Design - Develop - Deploy

Slide 32

Slide 32 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 33

Slide 33 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 34

Slide 34 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 35

Slide 35 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 36

Slide 36 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 37

Slide 37 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 38

Slide 38 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 39

Slide 39 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 40

Slide 40 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 41

Slide 41 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 42

Slide 42 text

Data acquisition Model Deployment Data Cleaning Feature Engineering Model Validation Model Monitoring Model Selection Model Testing Model Training Hyper parameter tuning

Slide 43

Slide 43 text

It’s time for exploring some FUN applications

Slide 44

Slide 44 text

Classification on Custom Data https://teachablemachine.withgoogle.com/train

Slide 45

Slide 45 text

Sketch-RNN https://magenta.tensorflow.org/assets/sketch_rnn_demo/index.html

Slide 46

Slide 46 text

From where the data came Quick Draw: https://quickdraw.withgoogle.com/

Slide 47

Slide 47 text

RUNN = 🏃Run + 🤖RNN http://vibertthio.com/runn/

Slide 48

Slide 48 text

Experiments with Google https://experiments.withgoogle.com/collection/ai

Slide 49

Slide 49 text

The project Magenta https://magenta.tensorflow.org/

Slide 50

Slide 50 text

https://developers.google.com/machine-learning/

Slide 51

Slide 51 text

No content

Slide 52

Slide 52 text

Let's connect! @CharmiChokshi Charmi Chokshi @charmichokshi coda.io/@charmi-chokshi/hi

Slide 53

Slide 53 text

Thanks! Questions? Download the deck