Machine Learning (ML) is being used by you dozen of time a day even without knowing about it. Having you ever realized each time you say Siri, Alexa or Google you are using ML. The Google search that you are doing in your computers or phones are using it. In this post we will discuss in detail about what is Machine Learning, how it works, its various types, applications, advantages and disadvantages.
Table of Contents
- 1 What is Machine Learning
- 2 How Machine Learning Works
- 3 Types of Machine Learning
- 4 Applications of Machine Learning
- 5 Advantages of Machine Learning
- 6 Disadvantages of Machine Learning
- 7 Conclusion
What is Machine Learning
Machine Learning is an application of Artificial Intelligence that provides systems the ability to automatically learn, predicts and improves from experience without being explicitly programmed.
Fig. 1 – Introduction of Machine Learning
It is a thoroughly algorithm driven study that makes computers, devices, software capable of learning on the basis of their own previous experience and improve the performance of a task. It also gives machines/software ability to analyze, predict and sort huge amount of data. The process of learning starts with data, instructions and observations to make better decisions in the future.
How Machine Learning Works
Let’s take an example to understand it in simple words. Assume we have data that contains pictures of different vegetables and we want machine learning to separate them.
Here’s how it works:
- First we provide data to the system.
- The system goes through the entire data and analyzes it to find patterns based on sizes, shapes, colors etc.
- Now after figuring out patterns, the system takes decisions and starts separating items.
- Once the work is done, the system learns from the result. If any vegetable type is wrong, it will make sure it does not happen again.
Fig. 2 – How Machine Learning Works
It works by building ‘smart algorithms’ and present the computer with ‘enough’ real world examples of the environment so that when computer sees ‘similar data’, it knows what to do.
Types of Machine Learning
Machine Learning can be sub-categorized to three types:
- Supervised Learning
- Unsupervised Learning
- Reinforcement Learning
Fig. 3 – Types of Machine Learning
Supervised Machine Learning
It trains a model on known input and output data so that it can predict future outputs. It usually predicts future events by using past learned or stored data using labeled examples. For example – Identifying a mail as spam or promotion.
Unsupervised Machine Learning
It finds hidden patterns or intrinsic structures in input data. Unsupervised ML algorithms are used when the information is neither classified nor labeled. For example opening of different blogs or website with the same search results.
Reinforcement Machine Learning
It builds a model that makes predictions based on proof in the presence of uncertainty.
Applications of Machine Learning
Following are some examples of ML Applications:
Virtual Personal Assistants
Virtual personal assistants Siri, Alexa, Google Now assist you in finding information when asked over voice and do the tasks assigned to be done.
Web Search Engine
One of the main reasons why search engines work so well is because the system has learnt how to rank pages through a complex learning algorithm.
Depending on your past searches, the Amazon or Facebook algorithms learn what sort of items you are looking for.
Machine Learning algorithms are used to analyze old stock data and predict the future values.
Photo Tagging Applications
In photo tagging applications, like Facebook, the ability to tag friends makes it even more interesting and exciting. All of this is possible because of a face recognition algorithm that runs behind the application.
Our mail agent like Gmail or Hotmail does a lot of hard work for us in classifying the mails and moving the spam mails to spam folder. The only possible way by which it is achieved is by the use of spam classifier running in the back end of mail application.
Predictions During Travelling
Machine Learning predicts which route has less congestion at that particular time by using traffic predictions. Not only this it will also help you in booking a cab from your place by applications like Uber and Ola.
Surveillance of Videos
A single person monitoring many video cameras is certainly a hard job to be done and boring too. This is why the idea of training computers to do this job is implemented using ML.
Fig. 4 – Applications of Machine Learning
Advantages of Machine Learning
The advantages of ML include:
Easily Identifies Trends and Patterns
It can process humongous volumes of data and discover specific patterns and trends. For example, for an e-commerce website like Amazon, it functions to understand the browsing histories and purchase behaviors of its users to showcase the matching items, deals, and reminders relevant to them.
No Human Intervention Needed (Automation)
With ML, you don’t need to babysit your project on every step of the way. As it means providing the machines the ability to learn, it lets them make predictions and also improve the algorithms on their own.
As Machine Learning algorithms gain experience, they keep improving in accuracy and efficiency. This lets them make better decisions.
Handling Multi-Dimensional and Multi-Variety Data
Its algorithms are excellent at handling data that are multi-dimensional and multi-variety, and they can do this in dynamic or random uncertain environments.
It has many wide applications such as banking, financial sector, healthcare, retail, publishing, etc.
Fig. 5 – Advantages and Disadvantages of ML
Disadvantages of Machine Learning
No matter how many advantages it has, it isn’t perfect yet. The following are its drawbacks:
It needs huge data to train on, and these should be unbiased and of good quality. There can also be times where the algorithm must wait for new data to be generated and fetched.
Time and Resources
It needs much time to let the algorithms adapt, learn and develop in order to fulfill their purpose with a considerable amount of accuracy and relevancy. It also needs huge resources to function.
Interpretation of Results
Another major challenge is the ability to accurately explain results generated by the algorithms. One must be careful in choosing the algorithms for their purpose.
High Error Susceptibility
It is autonomous but highly susceptible to errors. If you train an algorithm with data sets small enough to not be inclusive, you end up with biased predictions coming from a biased training set. This leads to irrelevant advertisements being displayed to customers. In the case of Machine Learning, such blunders can set off a chain of errors that can go undetected for long periods of time. And when they do get noticed, it takes quite some time to recognize the source of the issue, and even longer to correct it.
Fig. 6 – Machine Learning is Ever Improving
Machine Learning might have its very own disadvantages but than what doesn’t! In today’s scenario, where life is fast and human dependency on machines is ever increasing, it sure is a boon and with time it will evolve even more turning out to be full proof and maybe perfect!