Machine Learning(ML) – Introduction to Basics

Machine Learning (ML) – It is the hottest subject of today’s time. DataScientist is the sexiest job of today but the implementation of these buzzwords in real business is missing big time. The real need is to clarify, demonstrate, extract real values and reap rewards. “Machine Learning” sounds like a gold mine to many businesses, especially for the companies which are actually data factories i.e. social media platforms.

Machine Learning involves feature selection and feature engineering and encompasses a variety of algorithmic techniques. Techniques like deep learning and neural network etc. The deeper it gets, it starts hinting at the simulation of the human brain though and might get complicated for some.

---

Machine Learning(ML) – Basic Terminologies in Context

---

Machine Learning and AI

Artificial intelligence and machine learning are used interchangeably often but they are not the same. Machine learning is one of the most active areas and a way to achieve AI. Why ML is so good today; for this, there are a couple of reasons below but not limited to though.

• The explosion of big data
• Hunger for new business and revenue streams in this business shrinking times
• Advancements in machine learning algorithms
• Development of extremely powerful machine with high capacity & faster computing ability
• Storage capacity

Today’s machines are learning and performing tasks; that was only done by humans in the past like making better judgment, decisions, playing games, etc. This is possible because machines can now analyse and read through patterns and remember learnings for future use. Today the major problem is to find resources that are skilled enough to demonstrate & differentiate their learning from university & PhD books in real business rather than just arguing on social media with others.

Machine learning should be treated as a culture in an organisation where business teams, managers and executives should have some basic knowledge of this technology. To achieve this as a culture, there have to be continuous programs and roadshows for them. There are many courses which are designed for students, employees with little or no experience, managers, professionals and executives to give them a better understanding of how to harness this magnificent technology in their business.

Demystifying The Jargons

Artificial Intelligence (AI) and Machine learning (ML) are new buzzwords, which are commonly used by professionals and organizations. It is believed that these new technologies are going to impact every aspect of our lives and will open endless possibilities. AI has gone far ahead from the old-school thought of “AI is a bunch of if-else statements”.

So it’s very important to understand, how these new technologies/techniques are going to change our world. People who will not adopt these fast-changing technologies will be left behind. You all are welcome to learn with me at the speed and comfort of your own. Welcome to the machine learning journey let’s learn together. Sadly machine learning methods of today are mostly based on supervised machine learning (a type of machine learning), which means we still have a long way to go.

Make a note of the fact, if you really want to learn and have a good, strong and commanding attitude over machine learning then you can’t run away from mathematics, probability, linear algebra, statistics, decision theory, algorithms flow and a bit of neuroscience understanding at very high.

Machine Learning and Artificial Intelligence use them as underline foundations below 3 pillars i.e. computing power, data, and algorithms. ML is one of the technology/techniques under the umbrella of AI.

AI is used in many fields today some of them are Auto Pilot in Airplanes, Natural Language Processing, Robotics, Computer Vision, info-security and fraud detections.

What is Machine Learning?

Before we embark on our journey of defining machine learning, let’s understand it a bit. Machine Learning (ML) term got coined in the 1950s but it remained largely confined to academia then. Even to date, disappointingly machine learning is still treated as a complex subject and more accessible to developers only.  For more details on timelines & history of machine learning, please refer to this post – The Exciting Evolution of Machine Learning

We can call AI a colourful bundle of new techniques & emerging technologies. These technologies are available for everyone and anyone. No one has any proprietary claim of intellectual property etc on it. Machine learning is a subset of AI or one of the useful technology out of the AI bundle which help in achieving AI. Some experts define Artificial Intelligence as any technology, which enables us to demonstrate human-like intelligence. The current state of AI (Narrow AI) is based on 3 foundation pillars i.e.

• Data (Labels, knowledge, and information)
• Computing power (Brain or neural network)
• Algorithms (Logic and Experience)

AILabPage defines Machine Learning in one simple line as below

A focal point where business, data, and emerging technologies meets experience, and decide to work together.

The primary goal of machine learning (ml) is to build an automated data model for analytical reasons. The objective behind the goal is to build a system that learns from the data based on the applied algorithm. The output can be obtained by mapping output to input or detecting patterns/structure or learning by reward/punishment method.

The need for today is a simple & easily accessible machine learning cloud-based service i.e. “MLaaS – Machine Learning as a Service” for every individual and business at a low cost. The whole game revolves around building data models for analytical reasons and to find/see, what human eyes can’t see in its state then.

Types of Machine Learning

Machine Learning is classified into four categories at a high level depending on the nature of the learning and learning system.  Somehow I find it difficult to accept Semi-supervised learning as a separate class. The type of machine learning are

1. Supervised learning: Supervised learning gets labelled inputs and their desired outputs. The goal is to learn a general rule to map inputs to the output.
2. Unsupervised learning: The machine gets inputs without desired outputs, the goal is to find structure or hidden patterns for useful information in inputs. Extremely useful when we have a very large amount of unlabeled data.
3. Reinforcement learning: In this algorithm interacts with a dynamic environment, and it must perform a certain goal without a guide or teacher.
4. Semi-supervised Learning: This type of ml i.e. semi-supervised algorithms are the best candidates for model building in the absence of labels for some data. So if data is a mix of labels and non-labels then this can be the answer. Typically a small amount of labelled data with a large amount of unlabeled data is used here.

The approach of developing an ML model is to learn from data inputs with an idea as “What has happened”. Evaluating and optimising different model results remains the focus here. As of date Machine Learning is widely used in data analytics as a method to develop algorithms for making predictions on data as output. It has a high dependency on probability, statistics, and linear algebra.

Difference Between AI and Machine Learning

Artificial Intelligence (AI) and Machine Learning (ML) are commonly used interchangeably.  Both words do not mean the same thing although they are related. Machine Learning is one of the ways/methods to achieve AI / near AI. As mentioned before AI is an exceptionally wide umbrella kind of scenario.

AI is one of the biggest transformation measures of technology in today’s time. What we know machine learning is only one of the items in this AI bundle. We use these technologies knowingly or unknowingly on an almost daily basis. Examples are many of these including intelligent assistants making us more productive.

The AI definition is still not fixed or stable its like a moving target and keeps changing/evolving. As we achieve little more advancement we change the definition.  Surprisingly today’s AI might become the calculator technology of the past, remember what people use to say when the calculator came out for mathematical calculation.

It is very important to understand the difference between the two words and they necessarily do not solve the same problem. Depending upon the problem we are trying to solve we can decide to use ML or some other technique out of the AI bundle. Machine Learning is a technique out of the AI bundle to achieve artificial intelligence. Let’s just say AI generally defines & demonstrates creativity in the following traits:

• Planning & Predicting
• Learning & Adopting
• Reasoning & Logic Building
• Problem Solving & Avoiding
• Knowledge Representation
• Perception & Reasoning
• Motion Detection and Manipulation

Machine Learning helps to implement all the above by using correct methods and algorithms with correct data sets. In earlier times it was believed that human intelligence can be precisely described, and machines can simulate it with AI. Before the machine starts attempting simulation, it needs to do learning with lots of data.

The real fact is artificial intelligence is more about people, not just machines. Technology and non-technology companies are now investing and bringing out the real and materialistic values of Artificial Intelligence to the real world. Machine Learning is (mostly) a mathematics-specific AI technique for classification, regression, and clustering.

Now AI has started delivering values already. Using the contemporary view of computing exemplified by recent models and results from non-uniform complexity theory has proven the fact.

Algorithms in Machine Learning

Algorithms in Machine Learning (ML) borrow principles from computer science. There are many algorithms for the same purpose because no single algorithm can work best for every problem,

• How does youtube suggest you videos?
• Whether an email is a spam or not spam?
• Sentiment analytics to tell if a person is happy or sad?
• Which face has what name;  How facebook knows?
• How google map is able to find the best & fastest route between the office and home?

All the above question has one common answer “Algorithms”. The different problems will have different algorithms (solutions). ML instructs an algorithm to learn for itself by analysing data. The more data any algorithm processes, the smarter it becomes. Some of the popular machine learning algorithms are as below.

• Linear Regression – Simple Linear Regression.
•  Logistic Regression – The outcome variable is binary or dichotomous.
• Linear Discriminant Analysis –  A generalization of Fisher’s linear discriminant.
• Classification and Regression  Tree – Algorithm for predictive modeling
• Naive Bayes’ – A family of simple probabilistic classifiers.
•  K-Nearest Neighbours – Predict the numerical target based on a similarity measure.
• Learning Vector Quantization-  Representation of large amounts of data by (few) prototype vectors.
• Support Vector Machines- Employed for both classification and regression purposes.
• Bagging and Random Forest- Takes a bunch of bootstrapped samples from the original dataset.
• Boosting and AdaBoost-  Boosting is a machine learning ensemble meta-algorithm for primarily reducing bias.
• Q–learning is a model-free reinforcement learning technique.

Off-course each algorithm is used in the correct environment and usually doesn’t change the parent i.e. type of algorithm. Refer to the post “Machine Learning Algorithms“ for more details on machine learning algorithms and their use. Machine Learning entirely depends upon algorithms of two kinds i.e. Learning style and Symmetry & similarity.

Machine Learning Process

The process starts with data collection, cleaning and then training. Once training is completed by the data model it now needs to be tested with remaining data lying unused. The performance metrics showcase any gauge with the real world. This process can be iterative and in each iteration, performance parameters can be fine-tuned.

ML instructs an algorithm to learn for itself by analysing data. Algorithms here learn a mapping of input to output and detection of patterns or rewards. The more data it processes, the smarter the algorithm gets.

In other words, Machine learning algorithms “learn” from the observations. When exposed to more observations, the algorithm improves its predictive performance. The most basic rule and meaningful point to understand anything about machine learning are the quality of data it gets served with. So this means it’s purely driven by data sets to provide answers. It follows the GIGO method. The set of algorithms is as important as the underlying data sets.

Statistics and Machine Learning

The summation of Data, Mathematics, Statistics, Algorithms and Computing power can be called Machine Learning. It can be equated to magic but can be compared with futurology, which is the study of postulating possible, probable, and preferable futures. The world views and myths that underlie machine learning.

We have a classic example from linear regression (LR) which was developed in the field of statistics. Linear regression is studied as a model for understanding the relationship between input and output numerical variables. Now it has been borrowed by machine learning algorithms. It is both a statistical algorithm and a machine learning algorithm.

There are some methodological differences between machine learning and statistics but those really don’t divorce them. The difference between the two is that machine learning emphasizes optimization and performance, while others are concerned about sample, population, and hypothesis. Machine learning is more concerned with making predictions, even if the prediction cannot be explained well.

How to Lie with Statistics is a book written by Darrell Huff in 1954 presenting an introduction to statistics for the general reader. He was not a statistician but a journalist who wrote many “how-to” articles as a freelancer.

It is not magic, using familiar tools such as MS Excel, Python, R and machine learning cloud services from Azure and Amazon web services can clear this perception. This is a super cool and simplest subject you can ever encounter nothing to remember by heart.

Some of the uses for Machine Learning

Supervised Learning is becoming a good friend for marketing businesses in particular. For example how much money will we make by spending more dollars on digital advertising? Or even making small predictions for stock markets i.e. What’s going to happen to the stock market tomorrow?

Machine learning techniques are accelerating almost daily to bring good value to the businesses of today. It is revolutionizing the way we do our business and what should be done to improve upon it. ML develops its own encompassing strategy from the experience it comes across over the period.

Sir Andrew Ng says AI is a new form of electricity so, in my opinion, ML is no different; This has been one of the most active and rewarding areas of research due to its widespread use in many areas. Some of the major breakthroughs in ML include Natural Language Processing, and DeepLearning (GAN’s the most recent one).

Points to Note:

All credits if any remain on the original contributor only. We have covered all basics of machine Learning. Machine Learning is all about data, computing power and algorithms to look for information. In the previous post, we covered Generative Adversarial Networks. A family of artificial neural networks.

Books + Other readings Referred

• Research through open internet, news portals, white papers and imparted knowledge via live conferences & lectures.
• Lab and hands-on experience of  @AILabPage (Self-taught learners group) members.
• Our new member of AILabPage – Navdeep Kapur has contributed to this blog post.

Feedback & Further Question

Do you have any questions about Supervised Learning or Machine Learning? Leave a comment or ask your question via email. Will try my best to answer it.

Conclusion –  I particularly think that getting to know the types of machine learning algorithms actually helps to see a somewhat clear picture. The answer to the question “What machine learning algorithm should I use?” is always “It depends.” It depends on the size, quality, and nature of the data. Also, what is the objective/motive data torturing? As more we torture data more useful information comes out.

It depends on how the math of the algorithm was translated into instructions for the computer you are using. And it depends on how much time you have. To us, at AILabPage we say machine learning is crystal clear and ice cream eating task. It is not only for PhDs aspirants but it’s for you, us and everyone.

============================ About the Author =======================

Read about Author at : About Me

Thank you all, for spending your time reading this post. Please share your opinion / comments / critics / agreements or disagreement. Remark for more details about posts, subjects and relevance please read the disclaimer.

FacebookPage    ContactMe      Twitter

====================================================================