Types of Data Science Problems

Supervised Learning: When we know the labels in our dataset, when we have some idea about the dataset and what we are solving for. Ex: Find if a tumor is malignant or benign based on properties we know of.

 There are two types of supervised learning algorithms:
 1. Regression: When we want to find something continuous: Like price or height. 
 Ex : Given a dataset, find the price of an item, given the properties of the item.

 2. Classification: When we want to find if our target falls into discrete categories.
 Ex. Given a dataset, find if and item is over $5 or less then $5

Three(4?) Base Classes of Models (Supervised)

1. Linear Model (Support Vector Machine, Neural Network, Logistic Regression, Linear Regression )
2. Instance Based Model (K Nearest Neighbour)
3. Tree-based Model (Random Forrest, Decision Tree, XG Boost(any boost))
4. Bayesian Model

Unsupervised Learning: When we do not know the labels from our dataset. Sometimes, even if we have labels we could do some unsupervised learning before moving on to supervised learning. Ex: We are working on a military assignment where the variables are secret, but we have to find a relation and report our findings.