Artificial Intelligence versus Machine Learning versus Deep Learning

Reading Time: 7 minutes

The heading of this article comprises of the most searched keywords on Google today in the field of computer science. Every scholar, researcher has very high hopes and plans to solve the problems on our planet using these technologies.

What makes artificial intelligence, intelligent?

The answer to that question lies in the application under consideration. Broadly, we can define machine intelligence as:

  1. Robustness
  2. Ability to learn

Add the freedom to explore, and we would be on the brink of creating life in the form that we know today. These two qualities are very hard for traditional algorithms to attain, as the number of parameters under consideration is very high. Coming up with a mathematical model to solve the problem, would be a tedious task, and also, would be highly application-specific. For example, a model for recognizing apples and mangoes, would not be useful in the domain of recognizing faces, would it? On the contrary, when the algorithm understands the features and recognizes them, its intelligence and learning possibilities are enhanced significantly. It is no longer rote learning, but a quantitative understanding of the data through the features collected.

Why is this such a promising field?

  • Traditional computer science algorithms were application-specific and a lot of time was required to come up with an efficient algorithm for a specific application. Each application usually had different requirements and thus there was no one approach fits all method.
  • Artificial Intelligence, Machine Learning, and Deep Learning promise a subset of algorithms that are suitable for a variety of sub-tasks under a specific application domain. For example, object classification algorithms can be used to identify cats and dogs, but at the same time, the same model can be used to identify apples and mangoes. We will look into this concept a bit later.

How is artificial intelligence connected to machine learning?

  • Machine Learning is a subset of Artificial Intelligence, and is a domain, which deals with algorithms that improve over time. 
  • The definition of Machine Learning in terms of Tom Mitchell is as follows, “A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.”

Now that we have a basic understanding of artificial intelligence and machine learning, what is deep learning? Before we dive into deep learning, we need to understand the evolution of neural networks. 

How did the neural networks evolve?

–  Neural Networks are inspired partially by the neuron interconnection system in our brains. The notion of trial and error learning is emulated via neural networks.

– The neural networks were initially used to solve basic linear problems, until activation functions were introduced. Modelling the logical gates such as AND, OR, NAND, etc are linear problems. Linearity refers to the property of a single equation of a line being able to classify two classes.

Decision Boundaries for 'AND' and 'OR'

–  When one line isn’t sufficient to classify the two classes, that is, it is nonlinear, then a nonlinear equation is required. This is provided by the activation function, such as the sigmoid, ReLU, tanh, etc.

–  Activation functions introduce non-linearities and thus, the range of problems that can be solved using neural networks increases.

–  An application for the same would be modelling of the XOR gate, which requires two lines to define the boundaries of the two classes.

How do Neural Networks learn?

–  Neural networks mentioned above are one-shot learners. They only model the non-linearities in various models. 

–  Introduction of the back-propagation algorithm enhanced the learning capabilities of neural networks by a huge factor.

–  After the introduction of back-propagation algorithm and activation function, the neural network’s learning capabilities are enhanced.

What are some of the advantages of Back-Propagation Algorithm?

– Faster training

–  Efficient training

– Quantitative and qualitative measures of learning are available

– Works in real-time environment

For an in-depth discussion on neural networks, you can check this blog.

Where do neural networks fit into Deep Learning? Deep learning stems from the idea that complex problems with a large number of parameters can be solved by increasing the number of layers in an architecture. Thus, the first attempt to apply deep neural networks was recorded when the model LeNet was implemented. AlexNet, the ImageNet prodigy was inspired by LeNet, which won the ImageNet competition in 2012. 

Given an overview of all the three fields, we will go through some of the concepts and applications involved under each of these categories. This will enhance the understanding between the three subfields.

How can Artificial Intelligence be classified?

AI can be classified broadly into three categories:

  1. Analytical: Deals with only the intellect aspect of the brain
  2. Human-inspired: Deals with intellect, as well as emotional intelligence
  3. Humanized artificial intelligence: Understands intellect, emotions and the social structure

AI can be further classified based on the types of learning:

  1. Weak artificial intelligence: Machine programs work due to a well defined, highly efficient algorithm
  2. Strong artificial intelligence: The algorithms identify and learn the patterns, and thus least human involvement is involved after the design of the algorithm. Machine learning and deep learning are subsets of this field of AI.

What are the applications of AI?

  1. Applications: There are various applications of artificial intelligence systems. Let’s list a few:
    1. Maps services
    2. Recommendation Engines (Spotify, Netflix, Amazon)
    3. Robotics (Drones, Sophia the robot)
    4. Healthcare (Medical diagnosis, prognosis, precision surgery)
    5. Autonomous systems (Autopilot systems, self-driving cars)
    6. Drug discovery
    7. Stock market predictors

Having had a look at artificial intelligence, and its applications, we need to consider the following question.

How is machine learning different from artificial intelligence?

  1. Machine learning is a subset of strong artificial intelligence, where the algorithm’s performance improves over time. So many of the applications mentioned under artificial intelligence might overlap here too.
  2. Machine learning algorithms are classified into three types:
    1. Supervised learning
    2. Unsupervised learning
    3. Reinforcement learning

How are the various machine learning algorithms different?

The classification is based on the way the algorithm improves the performance measure P,  given the data-set.

Let’s consider a classroom experience, where the teacher is passing on the information to the students in a variety of ways like verbal recitations, writing on the board, activities, etc. What is the teacher doing? How is it different from a student sitting on his own and learning?

The answer to this question leads us to the difference between supervised and unsupervised learning algorithms. The role of the teacher is to teach and also tell the students what exactly they are learning, for example, in a physics class, the teacher tells the students what each equation means. 

On the contrary, if a student sits with the textbook which only has the equation in it, then, what the student makes out of the equation depends on her/his ability to comprehend mathematical equations. Alright, now let’s get technical. 

Supervised Learning algorithms work on data-sets which are labelled and thus, know what they are about to learn. In scenarios where we have an input variable x and an output variable G., The supervised learning algorithm learns the mapping between G and x as G=g(x), where g is the mapping function. Examples of supervised learning algorithms include all the classification algorithms, and the regression problems, like image classification, sentiment analysis, predicting the prices of used cars, etc. 

The main difference between classification and regression is that the output variable G is discrete in a classification problem, whereas it is continuous in a regression problem. Some major algorithms in this field of Machine Learning are Random forests, Support Vector Machines, Neural Networks (both shallow and deep). 

Unsupervised Machine Learning

  • In this scenario, the data is unlabeled and the main objective of the algorithm is to get as much relevant information out of the data-set given. The algorithm models the data-set or the distribution of the data-set and predicts its behavior given an input. 
  • Applications of unsupervised learning algorithms include clustering and association, like k-means clustering hierarchical ROCK clustering, and the famous association rule learning problems. 
  • Clustering essentially means the grouping of similar data points having similar properties together. 
  • Association learning is of interest majorly in data science, where the goal is to find what factors are inhibiting or boosting the sales of an iPhone let’s say.

Apart from this, there’s something called semi-supervised learning. The world where these two types of algorithms meet is the arena of semi-supervised learning, where the unsupervised learning algorithms are used as better feature extractors, and thus, as the rule of thumb goes for supervised learning algorithms, “the better the feature, the better the performance measure”. 

Reinforcement Learning

Reinforcement learning works on the concept of action and reward. It models our lives in a way, that is, makes incremental changes towards the optimum gradient by taking various possible actions. 

Consider an example of a drone. When a bird learns to fly, it starts flapping its wings, and slowly learns to fly. Whenever the bird falls down during a try, its reward is -1, and whenever it moves in the direction of the desired outcome, it gets a reward of +6. Hence, slowly it learns how to fly.

Applications of Machine Learning:

    1. Regression (Prediction)
    2. Classification(lesser number of classes, with less data)

 

  • Control Systems(Drones)

Structure of artificial neuron

Deep Learning:

Deep learning is part of the neural network based artificial intelligence. It’s roots lay in the concepts of optimization theory, and mathematical modelling of systems. The deep learning approach that works today did not work 10 years ago. The reason for this is:

  1. Enhanced, large and efficient datasets
  2. Huge rise in computational power capabilities

There are various types of architectures involved in deep learning. Each architecture has various basic units that make up the architecture. For example, machine translation systems have LSTM units. Classification models like Resnet have the residual block which repeats itself. The building blocks have desirable characteristics, and when repeated over time, result in positive results. 

How do you come up with your own architecture?

The answer would be practice and experience. Each layer, each change in the structure of the basic unit results in various desirable outputs. To learn how each concept affects the results, one should understand the basics, that can be found in our previous blogs.

There are various applications of deep learning as well:

  1. Translation systems (Google Translate)
  2. 3D Animation and effects (Avengers Endgame: Thanos)
  3. Navigation systems (Maps)
  4. Augmented Reality 
  5. Virtual Reality
  6. Healthcare (Disease prediction)
  7. Security (Detection of security breaches)

Overall, artificial intelligence comprises of machine learning and deep learning. machine learning is the go to set of algorithms, when data is limited, computational power is constrained and number of variables in the problem are less. deep learning deals with larger datasets, larger number of variables, and also requires tremendous power to run the models. We need to think about the brain at this moment. A small machine that can do all these tasks in less than 25 W of power. The future of these algorithms is the energy on which they shall sustain, and thus use of renewable energy is the need of the hour. 

 

Your Weekly Guide to Artificial Intelligence – September Part II – GL

Reading Time: 2 minutes

Artificial Intelligence is paving the way to the future, one breakthrough at a time. On the other hand, there are debates on the ‘pros and cons’ and the ‘perceived vs. actual intelligence’ of AI. Here are a few recent articles that highlight this notion. Read on to learn more.

A Breakthrough for A.I. Technology: Passing an 8th-Grade Science Test

The Allen Institute for Artificial Intelligence, a prominent lab in Seattle, unveiled a new system that passed the test with room to spare. It correctly answered more than 90 per cent of the questions on an eighth-grade science test and more than 80 per cent on a 12th-grade exam. The system, called Aristo, is an indication that in just the past several months researchers have made significant progress in developing A.I. that can understand languages and mimic the logic and decision-making of humans.

Artificial Intelligence Aid Fight Against Global Terrorism

Although terrorists have become skilled at manipulating the Internet and other new technologies, artificial intelligence or AI, is a powerful tool in the fight against them, a top UN counter-terrorism official said this week at a high-level conference on strengthening international cooperation against the scourge. Read more to know how AI technologies are being used to counter global terrorism.

AI is Not as Smart as You Think

‘Computers won’t cause the end of civilisation.’

Speaking at a recent artificial intelligence seminar, Dr Mariarosaria Taddeo, a research fellow at the Oxford Internet Institute, said AI will never think for itself. There is not a shred of proper research that supports the idea that AI can become sentient. This is a technology that behaves as if it were intelligent, but that is nothing to do with creating or deducing.

Top Highlights From The World Artificial Intelligence Conference

With the theme of “Intelligent Connectivity, Infinite Possibilities”, the World Artificial Intelligence Conference 2019 concluded recently in Shanghai. During the opening ceremony, Alibaba Group Chairman Jack Ma and Elon Musk, CEO of Tesla had a 45-minute debate on the impact of AI on human civilisation, future of work, consciousness and environment. The event saw innovative applications related to industrial ecology, AI urban application, autonomous driving and other cutting-edge technologies, and about 400 companies which participated.

How Artificial Intelligence is Creating Jobs in India, Not Just Stealing Them

There is a growing demand for data-labelling services that are “localised”- both linguistically and culturally relevant to India. From an opportunity point of view, there are about a lakh jobs posted on various portals currently. 

Happy Reading!

 

For more roundups on AI, watch this space!

If you are interested in upskilling with Artificial Intelligence, read more about Great Learning’s PG program in Artificial Intelligence and Machine Learning.

Difference Between Data Science, Machine Learning, and AI

Reading Time: 6 minutes

Even though the terms data science, machine learning, and artificial intelligence (AI) fall in the same domain and are connected to each other, they have their specific applications and meaning. There may be overlaps in these domains every now and then, but essentially, each of these three terms has unique uses of their own. 

We will start with the term Data Science, as it assumes the top-most position in the hierarchy of data-related technologies.

What is Data Science?

Data science is a broad field of study pertaining to data systems and processes, aimed at maintaining data sets and deriving meaning out of them. Data scientists use a combination of tools, applications, principles and algorithms to make sense of random data clusters. Since almost all kinds of organizations today are generating exponential amounts of data around the world, it becomes difficult to monitor and store this data. Data science focuses on data modelling and data warehousing to track the ever-growing dat set. 

The information extracted through data science applications can be used to guide business processes, which brings us to the next question- 

What are the Scopes of Data Science?

One of the domains that data science influences directly is business intelligence. Having said that, there are functions that are specific to each of these roles. Data scientists primarily deal with huge chunks of data to analyse the patterns, trends and more. These analysis applications formulate reports which are finally helpful in drawing inferences. A Business Intelligence expert picks up where a data scientist leaves – using data science reports to understand the data trends in any particular business field and presenting business forecasts and course of action based on these inferences. Interestingly, there’s also a related field which uses both data science and business intelligence applications- Business Analyst. A business analyst profile combines a little bit of both to help companies take data driven decisions.  

Data scientists analyse historical data according to various requirements, by applying different formats, namely:

    • Predictive causal analytics: Data scientists use this model to derive business forecasts. The predictive model showcases the outcomes of various business actions in measurable terms. This can be an effective model for businesses trying to understand the future of any new business move.  
    • Prescriptive Analysis: This kind of analysis helps businesses set their goals by prescribing the actions which are most likely to succeed. Prescriptive analysis uses the inferences from the predictive model and helps businesses by suggesting the best ways to achieve those goals.

Data science uses a wide array of data-oriented technologies including SQL, Python, R, and Hadoop, etc. However, it also makes extensive use of statistical analysis, data visualization, distributed architecture, and more to extract meaning out of sets of data.

Data scientists are skilled professionals whose expertise allows them to quickly switch roles at any point in the lifecycle of data science projects. They can work with AI and machine learning with equal ease. In fact data scientists need machine learning skills for specific requirements like:

  • Machine Learning for Predictive Reporting: Data scientists use machine learning algorithms to study transactional data to make valuable predictions. Also known as supervised learning, this model can be implemented to suggest the most effective courses of action for any company. 
  • Machine Learning for Pattern Discovery: Pattern discovery is important for businesses to set parameters in various data reports and the way to do that is through machine learning. This is basically unsupervised learning where there are no pre-decided parameters. The most popular algorithm used for pattern discovery is Clustering.

Sidetrade, a leading company in the domain of data-science, realized, early-on, the data exploitation challenges its clients faced and immediately set up a dedicated Data Scientist team to work with its Product Managers, aptly put it:

“Data Scientists, of course, have to work closely with IT development teams to guarantee the usability of any solution once it’s in production”

Jean-Cyril Schütterlé VP Product & Data Science, Sidetrade Group

Read Also: Artificial Intelligence and The Human Mind: When will they meet?

What is Artificial Intelligence?

AI, a rather hackneyed tech term that is used frequently in our popular culture – has come to be associated only with futuristic-looking robots and a machine-dominated world. However, in reality, Artificial Intelligence is far from that. 

Simply put, artificial intelligence aims at enabling machines to execute reasoning by replicating human intelligence. Since the main objective of AI processes is to teach machines from experience, feeding the right information and self-correction is crucial. AI experts rely on deep learning and natural language processing to help machines identify patterns and inferences.

What are the Scopes of Artificial Intelligence?

    • Automation is easy with AI: AI allows you to automate repetitive, high volume tasks by setting up reliable systems that run frequent applications.
    • Intelligent Products: AI can turn conventional products into smart commodities. AI applications when paired with conversational platforms, bots and other smart machines can result in improved technologies.
    • Progressive Learning: AI algorithms can train machines to perform any desired functions. The algorithms work as predictors and classifiers.
    • Analysing Data: Since machines learn from the data we feed them, analysing and identifying the right set of data becomes very important. Neural networking makes it easier to train machines.

Are machine learning and data science related?

Artificial Intelligence, much like data science is a wide field of applications, systems and more that aim at replicating human intelligence through machines. Artificial Intelligence represents an action planned feedback of perception.

Perception > Planning > Action > Feedback of Perception

Data Science uses different parts of this pattern or loop to solve specific problems. For instance, in the first step, i.e. Perception, data scientists try to identify patterns with the help of the data. Similarly, in the next step, i.e. planning, there are two aspects:

  • Finding all possible solutions
  • Finding the best solution among all solutions

Data science creates a system which interrelates both the aforementioned points and helps businesses move forward.

What is Machine Learning?

Machine Learning is a subsection of Artificial intelligence that devices means by which systems can automatically learn and improve from experience. This particular wing of AI aims at equipping machines with independent learning techniques so that they don’t have to be programmed to do so. 

Machine learning involves observing and studying data or experiences to identify patterns and set up a reasoning system based on the findings. The various components of machine learning includes:

  • Supervised machine learning: This model uses historical data to understand behaviour and formulate future forecasts. This kind of learning algorithms analyse any given training data set to draw inferences which can be applied to output values. Supervised learning parameters are crucial in mapping the input-output pair. 
  • Unsupervised machine learning: This type of ML algorithm does not use any classified or labelled parameters. It focuses on discovering hidden structures from unlabeled data to help systems infer a function properly. Algorithms with unsupervised learning can use both generative learning models and a retrieval-based approach. 
  • Semi-supervised machine learning: This model combines elements of supervised and unsupervised learning yet isn’t either of them. It works by using both labelled and unlabeled data to improve learning accuracy. Semi-supervised learning can be a cost-effective solution when labelling data turns out to be expensive. 
  • Reinforcement machine learning: This kind of learning doesn’t use any answer key to guide the execution of any function. The lack of training data results in learning from experience. The process of trial and error finally leads to long-term rewards.

Machine learning delivers accurate results derived through the analysis of massive data sets. Applying AI cognitive technologies to ML systems can result in the effective processing of data and information.

How are Data Science, Machine Learning, and AI Related?

Although it’s possible to explain machine learning by taking it as a standalone subject, it can best be understood in the context of its environment, i.e., the system it’s used within.

Simply put, machine learning is the link that connects Data Science and AI.

That is because it’s the process of learning from data over time. So, AI is the tool that helps data science get results and the solutions for specific problems. However, machine learning is what helps in achieving that goal.

Read Also: Difference Between Data Science & Business Analytics

A real-life example of this is Google’s Search Engine.

  • Google’s search engine is a product of data science
  • For instance, if a person types “best jackets in NY” on Google’s search engine, then the AI collects this information through machine learning
  • Now, as soon as the person writes these two words in the search tool “best place to buy,” the AI kicks in, and with predictive analysis completes the sentence as “best place to buy jackets in NY” which is the most probable suffix to the query that the user had in mind.

 

The diagram above is a helpful visual representation of the linkage between AI, Machine Learning and Data Science.

To be precise, Data Science covers AI, which includes machine learning. However, machine learning itself covers another sub-technology, which is deep learning.

Deep Learning is a form of machine learning but differs in the use of neural networks where we stimulate the function of a brain to a certain extent and use a 3D hierarchy in data to identify patterns that are much more useful.

Even though the areas of data science, machine learning and artificial intelligence overlap, their specific functionalities differ and have respective areas of application. The data science market has opened up several services and product industries, creating opportunities for experts in this domain. Upskilling in any of these areas will take your career a step ahead.

Click here to explore a career in Artificial Intelligence.