week6

Question 1

Support Vector Machine

Answer 1

• Another Linear Classifier
• SVMs use a boundary called a hyperplane to partition data into groups of similar values
• vector space-based machine learning method aiming to find a decision boundary between two classes that are maximally far from any point in the training data
• the goal of SMV is to create a flat boundary called a hyperplane which is a straight line that divides the space to create fairly homogenous partitions on either side
• the SMV learning combines aspects of both the instance-based nearest neighbor learning (lazy learning) classification using nearest neighbors and linear regression

Question 2

Application of SMV

Answer 2

• classification of microarray gene expression data in the field of bioinformatics to identify cancer or other genetic diseases
• text categorization such as identification of the language used in a document or the classification of documents by subject matter
• the detection of rare yet important events like combustion engine failure, security breaches, or earthquakes

Question 3

Collaborative Filtering Algorithm

Answer 3

• The technique used by Recommender Systems
• user-based filtering system
• if users A and B have purchased similar items in the past, the recommender system would have recommended items purchased by user B and to user A
• the similarity in behavior between two users is often computed by Cosine distance or Euclidean distance measure
• The lesser the cosine angle, the higher the similarity in behavior between the two users

Question 4

Content-Based Filtering Algorithm

Answer 4

• another algorithm used by recommender systems
• unlike collaborative filtering algorithm, content-based algorithms use features of the items such as genre, artist
• is user A has been buying Harry Potter books, it is likely that user A may purchase another fantasy book ‘the hobbit’
• one technique to computer similarity between two items is cosine similarity or Euclidean distance

Question 5

Feature Engineering

Answer 5

• process of extracting features from a raw dataset
• a term coined to give due importance to the domain knowledge required to select sets of features for machine learning algortihms
• as technology becomes more sophisticated, more datasets will be available
• but do we need all the features/variables of the dataset?

Question 6

Advantages of feature engineering

Answer 6

• improved predictive performance of the model
• faster and less complex machine learning process
• a better understanding of the underlying data relationships
• explainable and implementable machine learning models and solutions