Algorithm Definition: A step-by-step procedure for solving a problem or accomplishing some end especially by a computer.
An Algorithm But Not Machine Learning: Given a list of positive numbers, return the largest number on the list. Pretty straightforward procedure for solving for maximum. Is the 1st # larger than the 2nd, if yes, discard #2 and, repeat.
Simplest Machine Learning (ML) Algorithm: Given the 4 data points in the below chart, find the best-fitting straight line through the points so that the line minimized the distance all the points are from the line. The algorithm to solve this is called “linear regression” and the technical term that qualifies this as a machine learning algorithm is “optimizing a loss function”. Once the computer has learned the best fitting line based on a training data set, you can then predict Y given a new X value. For example, X is square feet of a house, Y is the price. ML algorithms get much more complex and there is a debate whether linear regression is really a ML algorithm or just statistics. The answer is, yes, according to this article
Click Here. At any rate, all machine learning algorithms, simple or complex, need good data, lots of it, to be trained.
Read next section “Show me the data!”.
