


Machine Learning Foundations:
A Case Study Approach
Machine Learning
Pipeline
(050416)

Case Study 1:
Predicting House Prices
(060416)

- Prediction of house sales price
- Looking at other houses and their house sales prices to inform the house valueof this house we're interested in.
- Looking at other features of the houses e. g. number of square feet
Case Study 2:
Sentiment Analysis
(060416)

- Looking at the text of
the review and the rating of the review, in order to understand what's the relationship for classification of the sentiment.
- Analyzinng the text of this
review in terms of how many
times it uses the word
Case Study 3:
Document Retrieval
(060416)
- In this case,data that we have is a
huge collection of possible articles
- intelligence we're deriving is an
article or a book that's of interest to
our reader
- findind structure in this data based
on groups of related articles

Case Study 4:
Product Recommendation
(060416)
- taking your past purchases +
other user's purchurse histories and
trying to use those to recommend
some set of other products
you might be interested in purchasing
- Using features of all customers and
features of products and trying to
compare and match those features
with your features


Case Study 5:
Visual Product Recommener
(060416)

- Here inputs are Images not Texts!
- Outputs are a set of results of shoes
that might also be of interest to a customer
and this customer want to be able to
search over those to purchase an Item
- In order to be able to go from an
image to a set of related images, we
need to have very detailed features
about that image to find other images
that are similar
- Deep Learning
- Each layer of the neural network provides
more and more descriptive features
Process of Machine Learning
(060416)

- Which task should be done do, e. g.
solving a sentiment analysis problem
- Which machine learning models should be
used, e. g. Support Vector Machine or
Regression
- Which methods should be used to
optimize the parameters of the model?
- An finally: the following questions:
Is this model really providing the
intelligence that I'm hoping for?
How do we measurethe quality of the system?
Going into depth in different Areas
(060416)

Which models or Technics should be used for specific Task?
Which Algorithms should be used for Optimizing the model?
Which key machine Learning Concepts should be used?
Regression
Case Study: Predictiong house prices
(060416)

Classification
Case Study: Analyzing Sentiment
(060416)

Clustering and Retrieval
Case Study: Finding Documents
(060416)

Matrix Factorization & Dimentiality Reduction
Case Study: Recommending Products
(060416)
