Machine learning is the study of algorithms and statistical models that computer systems use to perform a specific task without using explicit instructions. It is a branch of artificial intelligence.
Machine learning emphasises on automatic methods. It relies on patterns and inference. It gets machines to learn data and then make decisions on similar data without being explicitly programmed. It doesn't rely on rule-based programming.
Computers are able to see, hear and learn. Welcome to the future - Dave Waters
For all machine learning models, the efficiency is dependent on the quantity and quality of data. Large amount of data allows the “data to tell for itself,” instead of relying on assumptions and weak correlations.
There are three ways the machine learns:
a) Supervised learning - The algorithm is given a set of data and instructed what the expected output is. It can solve classification and/or regression problems.
Classification problem - When the output variable is a category, such as “red” or “blue” or “disease” and “no disease”.
Regression problem - When the output variable is a real value, such as “dollars” or “weight”.
In other words, a lot of data will be input to the system so that the machine can identify whether the conclusion that it has arrived at is right or wrong thanks to the labeling of data that is provided.
Examples: Random forest, decision trees, linear regression, native bayes
b) Unsupervised learning - Unsupervised learning algorithms are purposed to model structures, data distributions and workout the results themselves. It uncovers a hidden structure. It draws inferences from datasets consisting of input data without labeled responses.
In other words, there is no labeling and hence the machine has to figure out for itself if a certain decision was right or wrong thanks to the huge quantities of data that is fed to the system.
Examples: K-means, clustering
c) Reinforcement learning - Reinforcement learning concept revolves around agents taking actions based on the reward of their previous actions. It faces a game-like situation, where it employs trial and error to come up with a solution to the problem.
Reinforcement learning differs from the supervised learning in a way that in supervised learning the training data has the answer key with it so the model is trained with the correct answer itself whereas in reinforcement learning, there is no answer but the reinforcement agent decides what to do to perform the given task.
Examples: State–action–reward–state–action (SARSA)
Steps to build a machine learning model
Machine learning process involves the following steps:
Gather and clean data (sample) to represent large data (population) — this step can at times take the longest time
Learn and understand data to figure out trends and patterns
Build a model that understands the data and makes decisions on data
Feed the model 70%-80% of sample data. This set of data is known as Training Data
Validate the model with the rest of the data. This set of data is known as Test Data
Based on results, repeat the steps if required
A machine learning model answers several questions:
What is being learned?
How is the data being generated? In other words, where does it come from?
How is the data presented to the learner? For instance, does the learner see all the data at once, or only one example at a time?
What is the goal of learning in this model?
Skills required to build a machine learning model
Building a machine learning model requires a range of skills. It needs subject matter experts who already know about the problem. They understand data and can interpret data in an efficient manner. They already understand what features to use.
These experts often team up with technical experts. These experts build the intelligent algorithm model that learn from data and make calculated decisions. These experts decide on what model to use.
Once the algorithms are trained then these algorithms start forecasting and start making intelligent decisions. These decisions help analysts perform predictability analysis and help organisations improve productivity.
Also, tuning the model is a crucial step to improve accuracy of forecasted results. The objective is to find the optimum value for each parameter to improve the accuracy of the model.
Examples on Use of Machine Learning
Optical character recognition - Categorise images of handwritten characters by the letters represented
Face detection - Find faces in images (or indicate if a face is present)
Spam filtering - Identify email messages as spam or non-spam
Topic spotting - Categorise news articles (say) as to whether they are about politics, sports, entertainment, etc
Spoken language understanding - Within the context of a limited domain, determine the meaning of something uttered by a speaker to the extent that it can be classified into one of fixed set of categories
Medical diagnosis - Diagnose a patient as a sufferer or non-sufferer of a disease
Customer segmentation - Predict, for instance, which customers will respond to a particular promotion
Fraud detection - Identify credit card transactions (for instance) for instance may be fraudulent in nature
Weather prediction - Predict, for instance, whether it will rain tomorrow
Machine learning covers a broad area of uses and has many types of algorithm models. It is in effect learning. It is provoking shifts in culture in this new world of artificial intelligence. This article is a high-level introduction. In future posts, we will be writing more in-depth on the areas of machine learning.
Do you think machine learning can benefit your business? Share with us by leaving us a comment. If you require more information or assistance on machine learning, contact us. We want to be an extension of our clients. We help systematise process and decision-making. Subscribe to our newsletter for regular feeds.
Did you find this blog post helpful? Share the post! Have feedback or other ideas? We'd love to hear from you.
References
Princeton University, Machine Learning Course, https://www.cs.princeton.edu/courses/archive/spr08/cos511/scribe_notes/0204.pdf
Machine Learning Mastery, Supervised and Unsupervised Machine Learning Algorithms, https://machinelearningmastery.com/supervised-and-unsupervised-machine-learning-algorithms/, published on 16 March 2016
Towards Data Science, What's New in Deep Learning Research: Reducing Bias and Discrimination in Machine Learning Models with AI Fairness 360, https://towardsdatascience.com/whats-new-in-deep-learning-research-reducing-bias-and-discrimination-in-machine-learning-models-be116d3a571c, published on 24 September 2018