17 July
What is machine learning?

Machine Learning is nothing but application of Algorithms that learn from examples to solve a problem .The ingredients for Machine Learning techniques are computer science algorithms, Statistics and Artificial Intelligence.

When do you use machine learning?

  • A Pattern exists
  • Cannot pin down the mathematical relationship
  • We have data on it

Hence you seek the help of computer Algorithms to solve the problem for you without human intervention. 

Components of Machine Learning:

Let’s look at it with an example: Playing a Checkers game

A Computer program that learns to play Checkers might improve its performance as it’s measured by its ability to win at the task involving playing checkers games, through experience obtained by playing the games against itself

  • Task T: playing Checkers
  • Performance Measure P: %Games won against opponents
  • Training Experience E: Playing practice Game with itself

Types of Machine Learning Problems: 
Supervised Learning: 

Input data is labelled and you know what you are looking for. We build a model based on the known data (training data) and then the equation/algorithm of the model is tested on other datasets. The model “learns by example” and then predicts the output values. Example: Classification & Regression Problems

  • Classification: You are trying to model the differences or similarities between groups. For example fraud/non-fraud, default/non-default. [fraud,fraud,non-fraud, non-fraud,fraud,fraud]
  • Regression: Input Data is a known value with a real value (continuous set of points) rather than a group. For Example to try and predict the Monthly Sales of a company.

Unsupervised Learning: 

Input data is not labelled and you might not have an idea of what you are looking for. A model is prepared by finding patterns present in the input data.

Example: association rule learning and clustering

  • Clustering: Data is divided into groups based on similar characteristics and other measures of the data (based on distance between points). For Example: For a Retail Shop, classify groups of customers as High spenders, Medium Spenders and Low Spenders.You can group them based on factors such as Past Spend, Seasonal Spend, Demographics of the customer (Age, Income, etc.).

  • Association Rule Learning : Finding Rules between data points For Example for a Retail shop , if a Customer buys product A and you find a pattern that he also definitely buys Product B .And this relationship is observed for 70% of the times .Then based on this you can recommend product B to rest of the customers who buy Product A. You build association rules for each and every combination of purchases.

Reinforcement Learning (RL):

Enables an agent to learn from experience (in form of reward and punishment for explorative actions) and adapt to new situations, without a teacher. RL is an ideal learning technique for these data mining scenarios, because it fits the agent paradigm of continuous sensing and acting, and the RL agent is able to learn to make decisions on the sampling of the environment which provides the data. Nevertheless, RL still suffers from scalability problems, which have prevented its successful use in many complex real-world domains. The more complex the tasks, the longer it takes a reinforcement learning algorithm to converge to a good solution. Example problems are systems and robot control. Example Algorithms are Neural Networks.

Interested in learning about other Analytics and Big Data tools and techniques? Click on our course links and explore more.
Jigsaw’s Data Science with SAS Course – click here.
Jigsaw’s Data Science with R Course – click here.
Jigsaw’s Big Data Course – click here.

 
 

SHARE
share

Are you ready to build your own career?