Machine learning workshop, session 5. - Support Vector Machine - Real Application - People to follow

1. Machine Learning
Lunch & Learn - Session 5
Luis Borbon
13/07/2017

2. Table of contents
1. Recap
2. Support Vector Machine
3. Real Application
4. People to follow

3. Recap

4. Induction and deduction
Induction refers to learning general concepts
from specific examples which is exactly the
problem that supervised machine learning
problems aim to solve.
This is different from deduction that is the other
way around and seeks to learn specific concepts
from general rules.

5. Overfitting and Underfitting

6. Algorithms by Similarity (cont…)

7. Support Vector Machine (SVM)
It is a classification method. In this algorithm, we
plot each data item as a point in n-dimensional
space (where n is number of features you have)
with the value of each feature being the value of
a particular coordinate.

8. Support Vector Machine
We have a population composed of 50%-50% Males and Females. Using a sample of this population, you
want to create some set of rules which will guide us the gender class for rest of the population.

9. Support Vector Machine
Find the boundary that separates classes by as
wide a margin as possible.
When the two classes can't be clearly separated,
the algorithms find the best boundary they can.
Because it makes this linear approximation, it is
able to run fairly quickly. Where it really shines is
with feature-intense data, like text or genomic. In
these cases SVMs are able to separate classes
more quickly and with less overfitting than most
other algorithms, in addition to requiring only a
modest amount of memory.

10. Support Vector Machine

11. Support Vector Machine
Not a clear frontier?

12. Support Vector Machine
Data that is not linearly separable?
http://efavdb.com/svm-classification/

13. Support Vector Machine
● Pros:
○ It works really well with clear margin of separation
○ It is effective in high dimensional spaces.
○ It is effective in cases where number of dimensions is greater than the number of samples.
○ It uses a subset of training points in the decision function (called support vectors), so it is also memory
efficient.
● Cons:
○ It doesn’t perform well, when we have large data set because the required training time is higher
○ It also doesn’t perform very well, when the data set has more noise i.e. target classes are overlapping
○ SVM doesn’t directly provide probability estimates, these are calculated using an expensive five-fold
cross-validation. It is related SVC method of Python scikit-learn library.

14. Real Application

15. PopGun.ai
Popgun is using Deep Learning to create an
exciting new musical experience.
We are building an AI that can play music with
you, just like a professional musician.
We think this AI will become an essential
learning and creative tool for musicians
worldwide.

16. People to Follow

17. Brand within the AI social media debate

18. Jensen Huang
Jensen Huang is a Taiwan-born American entrepreneur and
businessman.
He co-founded the graphics-processor company Nvidia and
serves as its president and CEO.
Huang graduated from Oregon State University before
moving to California.
● GTC 2017
https://youtu.be/WLq9zv3k5n0