personalized ml
TRANSCRIPT
![Page 1: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/1.jpg)
Introduction in Machine Learning - Personalizing
with ML
![Page 2: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/2.jpg)
What is Machine Learning?
“A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P, if its performance at tasks in T, as measured by P, improves with experience E.”
-- Tom Mitchell, Carnegie Mellon University
![Page 3: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/3.jpg)
A Spam Filter Application
E = The experience of classifying mail as spam or not spam.
T = The task of classifying emails (by subject, spam - not spam,
sender, ….), in our case whether it’s spam or not.
P = The probability that a mail is spam.
![Page 4: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/4.jpg)
So the goal in (nearly) all ML applications is...
For a function that predicts a quantity Y as , where epsilon is the random error we try to minimize the difference between the predicted and the real real value of Y:
![Page 5: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/5.jpg)
Getting our hands dirty with Linear Regression
http://bit.ly/28Zv65R
![Page 6: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/6.jpg)
Prediction vs Classification
Prediction: Quantitative answers, predict a value of a continuous variable (height, weight, mileage, Click-through Rate, …)
Classification: Qualitative answers, find what is the observation you are looking at (a dog or a cat, span or not spam, dog breed, ….)
![Page 7: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/7.jpg)
Classifying animals by their specie
![Page 9: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/9.jpg)
What is a neuron?
![Page 10: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/10.jpg)
How many of them we have in our heads?
![Page 12: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/12.jpg)
And the Homo Sapien’s answer: The artificial neuron!
![Page 14: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/14.jpg)
A neural network with multiple levels of features (intermediate processing layers)
What is Deep Learning
![Page 15: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/15.jpg)
We can also optimize the structure of every layer, the transfer function, the activation function φ, the thresholds θ and much more.
Deep Learning is not limited in number of layers
![Page 17: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/17.jpg)
Use Keras to train a network that classifies your photos against another person’s.
Homework
![Page 18: Personalized ml](https://reader031.vdocuments.us/reader031/viewer/2022030309/58f1cbc41a28abf9398b469f/html5/thumbnails/18.jpg)
If you want to learn more:
Thank you for your time and effort!