Supervised and Unsupervised Learning Flashcards
What is supervised learning?
Algos that are designed to learn by example and trained on well-labeled data and consists of an input object (typically a vector) and a desired output (supervisory signal)
What is the objective of leveraging a supervised learning model?
Predict the correct label for unseen data
Explain at a high level how a supervised learning model works?
It takes an input object which is typically a vector and a desired output (supervisory signal) and during the training phase, it searches for patterns that correlate with the desired output. After training is complete, it takes in unseen inputs and determines which label to classify it to
Supervised learning can be categorized into which subcategories?
Classification and Regression
How do classification algorithms work?
Assigns input values to classes or categories and an example would be determining if an email is spam or not spam
What are some of the popular classification algorithms?
linear classifiers, support vector machines, decision trees, k-nearest neighbors, and random forest.
What is a regression algorithm
predict continuous numbers, such as sales, income, and test scores and the goal is to find the relationship between dependent and independent variables
Examples of common regression algos?
Linear regression, lasso regression, and multivariate regression
What is unsupervised learning?
Focuses on datas features and does not use labeled data
Difference between supervised and unsupervised learning?
supervised learning uses input and outputs aka labels to train the model while unsupervised learning focuses on features
What is the goal of unsupervised learning?
Purpose is to analyze data and find important features in that data
What can unsupervised learning do that supervised cannot?
Find subgroups or hidden patterns within a dataset that a human observer might not pick up on
What are the two subcategories of unsupervised learning?
Clustering and Association
What is clustering?
process of grouping the given data into different clusters or groups
What are the components that go into clustering?
Classes will contain data points that are similar as possible to each other and as dissimilar as possible to data points in other clusters
What is the objective of clustering?
Helps to find underlying patterns within the data that may not be noticeable through a human observer
What is partitional clustering?
refers to a set of clustering algos where each data point in a data set can belong to only one cluster
What is hierarchical clustering?
finds clusters by a system of hierarchies; every data point can belong to multiple clusters.
What is association?
attempts to find relationships between different entities. market basket analysis is a good example of how someone that usually buys burgers and fries will usually buy a drink with those items
What are some use-cases for using unsupervised learning?
Airbnb - customer queries their requirements and airbnb learns those patterns and recommends stays and experiences which fall under the same group of cluster
