[MUSIC] As Emily discussed, we're gonna see machine learning through
the lens of a wide range of case studies in different areas that really
ground the concepts behind them. So, other machine learning classes
that you might take out there are really a laundry list
of algorithms and methods. So things like support vector machines and kernels and logistic ration and
networks and so on. And they're just like,
a laundry list of methods. And the problem with that approach is that
since you start from the algorithms you end up with really simplistic
use cases with the applications, they're really disconnected from reality. So, we're doing things very different
in this specialization, and we've done this for quite a while here. Emily and I created a course at
the University of Washington on machine learning at scale for big data. We pioneered this use case approach for
teaching machine learning. And in that course,
we saw a lot of positive feedback from folks really understanding
rounding the concepts. So we're going to start from
the use cases in the first course. And by starting from use cases, you're
really going to be able to grasp the key concepts and the techniques that allow
you to build, measure the quality and understand whether your intelligent
applications is working well or not. And in the end, you are going to build
a bunch of these intelligent applications. So to build such intelligent applications, you typically have to think about
what task am I going to do. I am going to solve a sentiment
analysis problem and what models, what machine learning models am I going to use,
and things like support vector machines or regression what methods when they use to
optimize the parameters of that model? And then I ask a question like is this
really providing the intelligence that I'm hoping for? How do we measure
the quality of that system? So in this specialization what
we're gonna do is defer the core pieces of how to describe a model and
optimize it to the follow on courses. And this first course is going to be
focused on helping us figure out what task we're trying to solve, what machine
learning methods make sense, and how to measure them. And with that, using the algorithms as
black boxes, we're going to be able to build a wide range of really
intelligent cool applications together. And we'll actually code them and build them and
demonstrate them in a wide range of ways. Now the following on courses, they're, it's going to be four of
those plus a capstone. They really go into depth
in different areas. So let me give you a few quick examples
of the kind of depth we're going to see throughout this specialization. So the regression course is going to
talk about various models of predicting a real value, so for example, a house
price from the features of the house. And we're going to discuss
linear regression techniques, we're going to discuss advanced techniques
like ridge regression and lasso that allow you to select what features are most
appropriate for your problem. We're going to talk about optimization
techniques like gradient descent and coordinate descent to optimize
the parameters of those models. As well as some key machine learning
concepts like loss functions, bias-variance tradeoffs, cross-validation. Things that you need to know to really
take this method and kind of improve them, develop them and
build applications with them. The second course on classification,
we're gonna build, for example, the sentiment analysis use case
that Emily talked about, and talk about more of those classifications. From linear classifiers to more
advanced things like linear regression, sorry, logistic regression,
support vector machines. But then add kernels and decision trees which allow you to deal
with non-linear complex features. We talked about optimization methods for
dealing with these techniques at scale and for building ensembles of them
something called boosting. And then the underlying concepts in
machine learning that really help you grasp classifier and scaled it up and
apply it to different methods. Now, in the next course,
we're gonna focus on clustering and retrieval, especially in
the context of documents. So we're gonna talk about basic
techniques like nearest neighbors as well as more advanced clustering
techniques, mixture of Gaussians, and even latent Dirichlet allocation can advance
text analysis clustering technique. We're gonna talk about the algorithms
that underpin these things and how to scale them up with
techniques like KD-trees and sampling and expectation maximization. Now the core concepts here are really
around how to scale these things up, how to measure the quality and
really how to write them as distributed algorithms using
techniques like map-reduce, which are implementing systems like
Hadoop that you might have learned about. So in the fourth course, you're actually
going to write some map-reduce code for distributed machine learning. Now in the final technical course we're
gonna focus on techniques of matrix factorization and dimensionality
reduction, which are widely applicable, but in particular for recommender systems,
for recommending products. So these are things like
collaborative filtering, matrix factorization, PCA, and
the underlying techniques for optimizing them, like coordinate descent,
Eigen decomposition, SVD. And then, a wide variety of whole machine
learning concepts that are really useful. Especially in the recommenders' domain. Like how to pick a diverse
set of recommendations and how to scale them up to large problems. Now the capstone is going to be really
exciting and towards the end of this module, I'm going to go back and tell
you quite a bit more about the capstone. But just to give you a little hint,
you're going to build something extremely cool that you can show
to all your friends, potential employers. And you'll see that it can build
a really smart intelligent application around recommenders,
the combined text data, image data, sentiment analysis, deep
learning, it's going to be really cool. [MUSIC]