Lets Understand AI with ease.
Note: This is an aggregated article from various blogs website. Aimed at helping people understand AI with ease.
Machine Learning (ML)
Deep Learning (DL)
AI, ML, and DL Difference and Timeline
Foundations of ML
What does Learning mean?
What is ML Model?
Types of Machine Learning
Supervised Learning
Unsupervised Learning
Reinforcement Learning
ML Classification and Use cases
Artificial Neural Networks & Deep Neural Networks
Future of AI
Artificial Intelligence (AI)
AI is the simulation of human intelligence processes by machines, especially computer systems. These processes include
- · Learning (the acquisition of information and rules for using the information)
- · Reasoning (using the rules to reach approximate or definite conclusions)
- · Self-correction.
Machine Learning (ML)
Machine learning is a type of AI that allows software applications to become more accurate in predicting outcomes without being explicitly programmed.
Deep Learning (DL)
A sub-field within ML that is based on algorithms for learning multiple levels of representation in order to model complex relationships among data. This is highly inspired by our understanding of how the human brain works.
AI, ML, and DL Difference and Timeline
AI, machine learning, and deep learning are not the same things. The easiest way to think of their relationship is to visualize them as concentric circles with AI — the idea that came first — the largest, then machine learning — which blossomed later, and finally deep learning — which is driving today’s AI explosion — fitting inside both.
In this document, we are going to have a detail look at ML and DL.
Foundations of ML
The foundational disciplines that contribute to the field of machine learning are:
- Statistics
- Mathematics
- Computation
What does Learning mean?
Learning can be visualized as a combination of Memorization and Generalization in the context of ML
· Memorization is the act of mastering and storing a pattern for future recollection.
· Ability to extrapolate using the principles Memorization to solve new examples is known as Generalization.
What is ML Model?
The “ML model” is the output generated when you train your “machine learning algorithm” with your training data-set. This can be visualized as a formula that generates predictions as output for the new data set. Below diagram explains how a ML model is generated (Final Hypothesis refers a ML model).
Types of Machine Learning
The three distinct blocks that characterize machine learning are:
- · Supervised learning
- · Unsupervised learning
- Reinforcement learning
Supervised Learning
The idea of supervised learning is to teach the computer using labeled data. The computer learns from data, the patterns that correspond to the desired output to generate a predictive ML model. Supervised learning can be further grouped into regression and classification problems.
A classification problem is when the output variable is a category, such as “Will a flight be overbooked”.
A regression problem is when the output variable is a real value, such as “Predictive booking count on given day and route”
Some popular examples of supervised machine learning algorithms are:
Unsupervised Learning
In unsupervised learning, there is no such guidance with labeled data, rather the computer attempts to determine the unknown structure of the data by grouping similar samples of a particular feature. Unsupervised learning can be further grouped into clustering and association.
Clustering is where you want to discover the inherent groupings in the data, such as “Segmentation of customers based on the different attributes”.
Association is a rule learning problem where you want to discover rules that describe large portions of your data, such as “People who traveled for Seashells also travels for Maldives”.
Some popular examples of unsupervised machine learning algorithms are:
· Apriori association
When we don’t have a labeled data but we expect an out of regression or classification problem a hybrid learning called Semi-Supervised Machine Learning is used. This is a combination of unsupervised learning and supervised learning.
Association is a rule learning problem where you want to discover rules that describe large portions of your data, such as “People who traveled for Seashells also travels for Maldives”.
Reinforcement Learning
The concept of reward and penalty is central to reinforcement learning. Here the computer interacts with an environment that is changing, its rewarded when the response from the environment is favorable and penalized when the response is unfavorable. The value function of the learning task determines the decision on what makes an action favorable or unfavorable.
Some popular examples of reinforcement machine learning algorithms are:
· Sarsa algorithm
ML Classification and Use cases
Details of various use cases with machine learning can be clearly visualized with the below info graphics.
Artificial Neural Networks & Deep Neural Networks
Artificial neural networks (ANNs) are computing systems inspired by the biological neural networks that constitute animal brains. It’s is based on a collection of connected units called artificial neurons. Each connection (synapse) between neurons can transmit a signal to another neuron. The receiving (postsynaptic) neuron can process the signal(s) and then signal downstream neurons connected to it.
A deep neural network (DNN) is an ANN with multiple hidden layers between the input and output layers. This has been introduced with the objective of moving Machine Learning closer to one of its original goals: Artificial Intelligence. A Neural Networks can be visualized similar to below infographics.
Neural Network Use Cases
Some classic use cases of Neural Network are
Some classic use cases of Neural Network are
· Automatic speech recognition
If we divide AI into three stages,
· The first stage is limited to only one functional area
· The second stage AI should be able to combine different narrow areas to perform tasks on a human skill level.
· The final stage is an intelligence, which surpasses the human capabilities.
Currently, we are in the transition from the first to the second stage.
References
Comments
Post a Comment