Astonishing Hierarchy of Machine Learning Needs – Artificial intelligence and machine learning are used interchangeably often but for 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. Machine Learning entirely depend upon algorithms of two kinds
- Learning style
- Symmetry & similarity
This post only depicts the high-level summary to gain simple, easy and layman idea. This blog post is not for PhD students or for experienced professionals in the industry but for sure it might just give them some smiling effects.
Machine Learning is the hottest subject of today’s time, DataScientist is the sexiest job of today but implementing these buzzwords in real life business is the most important need. The real need for today’s time and business is to clarify, demonstrate and extract real values to benefit everyone from this golden keyword “Machine Learning”. Why ML is so good today; for this, there are a couple of reasons like 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
As on date sadly most of the machine learning methods are based on supervised learning. Which means we still have a long long way to go. In Fintech domain we say #MachineLearning is the future (actually that future is now) of
#Ecommerce & #DataScientist will work as a batman for #FinTech & #InsureTech. Today’s machines are learning and performing tasks; that was only be done by humans in the past like making a better judgement, decisions, playing games etc.
Lets Define Machine Learning
Arthur Samuel coined in 1959. He called it a “field of study that gives computers the ability to learn without being explicitly programmed.”
Machine Learning is hottest subject of today’s time, DataScientist is the sexiest job of today but implementing these buzz words in real life business is most important need.
Machine Learning is a focal point where business needs and experience (Mathematics, Statistics & Algorithmic logic/thinking) meet emerging technology and decides to work together to put useful results on the table for real business.
Some History of Machine Learning
ML Arthur Samuel coined in 1959. Some of the timeline of machine learning. Machine learning has evolved from Artificial Intelligence Subset to Its own domain. It has reached an inflexion point – at least in terms of messaging. I remember in my school days as part of statistics class we were told something about AI and ML and we laughed then in the 1990s.
Although Machine Learning has now gained prominence owing to the exponential rate (1990 was the flying gear year) of data generation and technological advancements to support it has roots from the old days. Major discoveries, achievements, milestones and other major events are included in the picture here.
Types of Machine Learning
The approach of developing ML includes learning from data inputs based on “What has happened”. Evaluating and optimizing different model results remains focus here. As on date Machine Learning is widely used in data analytics as a method to develop algorithms for making predictions on data. It is related to probability, statistics, and linear algebra.
Machine Learning is classified into three categories at a high level depending on the nature of the learning and learning system.
- Supervised learning: Machine gets labelled inputs and their desired outputs. The goal is to learn a general rule to map inputs to the output.
- Unsupervised learning: Machine gets inputs without desired outputs, the goal is to find structure in inputs.
- Reinforcement learning: In this algorithm interacts with a dynamic environment, and it must perform a certain goal without a guide or teacher.
In a hypothetical situation or most of the time (At least from our personal experience), the amount of data anyone will find may look like the picture above. We are talking about the volume of the data. The volume of data for Supervised Learning is highest and for reinforcement learning its the lowest almost all the times.
Data Science & Machine Learning
Thanks to the statistics, machine learning became very famous in 1990s. The intersection of computer science and statistics gave birth to probabilistic approaches in AI. This shifted the field further toward data-driven approaches. Machine Learning is about the use and development of fancy learning algorithms. Data science is more about the extraction of knowledge (KDD) from data through algorithms to answer a particular question or solve particular problems
In large-scale data available systems, intelligent systems with best suitable algorithm analyze, detect patterns and learn to inform decision and information. Machine learning helps data science by making a provision for data analysis, data preparation and even decision making. The word learning in machine learning means that the algorithms depend on some data, used as a training set, to fine-tune some model or algorithm parameters.
Machine Learning is contained inside Data Science, every time an ML algorithm is used the process is as good as doing Data Science. As explained above ML algorithms learn from data which is the essential fuel of Data Science. Not 100% accurate or correct though but comfortably we can say Machine learning and statistics are part of data science.
Hierarchy of Machine Learning Needs
One thing we need to remember overdose of anything is bad and means bad. After doing everything or doing things in a rush there is no guarantee that machine learning and AI will improve anything. Installing the Ferrari engine in 4 x 2 ordinary car and expecting it to perform like Ferrari is surely not a good idea or work style. Similarly, the most advanced data analytics tools may simply get you to the outcome faster. But correctness is not guaranteed.
There is a need to collect correct, accurate, quality and relevant real-time data, that is organised, clean, tested and optimised (We may have to repeat this cyclic process many times though. The time to test machine learning and artificial intelligence solutions arrive after much-required efforts.
Machine Learning – An Independent Domain
Artificial Intelligence and Machine Learning are two very hot buzzwords right now, and often seem to be used interchangeably. In my best of knowledge and experience Machine Learning is growing at top gear and going make this statement true “ML is not AI and has its own independent identity”. AI is all about smart devices and applications but Machine Learning is all about learning and to gain accuracy. Maybe ML’s finished product can be called as AI.
Anyways our intention here is not to justify or show the difference between these 2 buzzwords but to prove Machine Learning is and going to be independent of AI. Machine Learning would grow out as matured enough technique (set of techniques) independently that it would not be needing AI as a superset of it.
Machine learning touches more of generalised AI rather than applied AI. ML not at all fit in definition if we say “replicating the process of human intelligence through algorithms” but certainly its AI. Machine Learning is nothing like actual intelligence, it is closer to a process of artificial model creation than anything else.
Machine Learning is not Magic
The summation of Data, Mathematics, Statistics, Algorithms and Computing power can be called as 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 worldviews and myths that underlie machine learning.
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.
At AILabPage we say this is crystal clear and ice cream eating the job. It is not only for PhDs aspirants but it’s for you, us and everyone. Some time back Microsoft started to add Machine Learning components to Azure. At that time people got overwhelmed and confused by the enormous number of technologies and jargons surrounding it. With Google announcing TensorFlow and Cloud ML followed by Amazon’s launch of its own Machine Learning service, it started to become very clear that ML is going to be the next big thing in the cloud.
Real Life Use Cases
Machine learning is the process of a machine attempting to accomplish a task, independent of human intervention, more efficiently and more effectively with every passing attempt i.e learning phase. At this point, AI- a machine which mimics the human mind, is still a pipe dream. In the middle, we have the meat of the pipeline, the model, which is the machine learning algorithm that learns to predict given input data.
ML a sub-set of artificial intelligence (shouting for independence and getting there slowly) where computer algorithms are used to autonomously learn from data and information. Machine Learning, however, has been a reality in our lives for quite some time on the other hand Deep learning is a subcategory of machine learning algorithms that use multi-layered neural networks to learn complex relationships between inputs and outputs.
Machine Learning to bring – Data Intelligence as a Service
In coming times data intelligence services will be the most eminent application that would be provisioning prototypes for security measures to truly fortify the DIaaS.
There are certain market dynamics which determine the growth of the data and its related analytics. That’s where Data Intelligence’s adaptive dynamics comes into play to assess the factors driving the organisation to adapt their existing, profitable lines of business. This helps them to stay relevant in the future of the rapidly evolving world and enormous helper for the Blue Ocean shift strategy.
Machine Learning use cases for Info-Security (#GDPR)
Machine Learning can help to automate more menial tasks previously carried out by IT or network security skilled teams who anyways are neither qualified nor skilled for this type of info-security especially in financial system data security.
In principle, machine learning can help businesses better analyze threats and respond to attacks and security incidents. The results of ML protecting data is astonishing and ML as info-security guard is much better than any human. Thusly, machine learning in info-security is a quickly developing & detecting anomalies in patterns.
I guess we have now done enough of drum beating for Machine Learning and all those learnings around it, its time to have some action and get this done on the ground for our real & daily life use.
Points to Note:
All credits if any remains on the original contributor only. We have now summarised our last four posts here to give a quick glimpse. You can find previous posts on Machine Learning – The Helicopter view, Supervised Machine Learning, Unsupervised Machine Learning and Reinforcement Learning links.
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.
- Machine Learning – An Introduction
- Machine Learning -A Probabilistic Perspective
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 – Traditional Machine Learning create train/ test splits of the data; possibly via cross-validation. Load ALL the training data into the main memory and Compute a model from the training data (may involve multiple passes. This what we have depicted here in the post. Machine Learning is a growing field that is used when searching the web, placing ads, credit scoring, stock trading and for many other applications. How machine learning uses computer algorithms to search for patterns in data? To use the data patterns to make decisions and predictions with real-world examples from healthcare involving genomics and preterm birth. To uncover hidden themes in large collections of documents using topic modelling. We will see the responses to these questions in subsequent blog posts.
============================ About the Author =======================
Read about Author at : About Me
Thank you all, for spending your time reading this post. Please share your feedback / comments / critics / agreements or disagreement. Remark for more details about posts, subjects and relevance please read the disclaimer.
Categories: Machine Learning